1 00:00:00,280 --> 00:00:02,400 Hi, and welcome to the Neil Ashton Podcast. 2 00:00:03,080 --> 00:00:05,920 In each episode, we explained some of the fascinating ways 3 00:00:05,920 --> 00:00:09,080 that science and engineering are changing the world around us. 4 00:00:09,800 --> 00:00:12,760 We talked to leading engineers from elite level sports like 5 00:00:12,840 --> 00:00:16,880 cycling in Formula One to some of the world's top academics to 6 00:00:16,880 --> 00:00:20,960 understand how fluid dynamics, machine learning, supercomputing 7 00:00:21,360 --> 00:00:23,000 are bringing in a new era discovery. 8 00:00:23,960 --> 00:00:27,080 We also hear some of their life stories, their career advice, 9 00:00:27,640 --> 00:00:30,080 the lessons they've learned on the way that I hope will be 10 00:00:30,080 --> 00:00:33,800 helpful to you too. So sit back and enjoy this 11 00:00:33,800 --> 00:00:41,120 episode. Hi, and welcome back to the Neil 12 00:00:41,120 --> 00:00:44,360 Ashton Podcast. So today I'm speaking with 13 00:00:44,480 --> 00:00:50,080 Professor Mike Giles, somebody who has had a enormous influence 14 00:00:50,160 --> 00:00:56,480 on the CFD community, but more broadly in, in maths, in high 15 00:00:56,480 --> 00:01:01,680 performance computing. And someone that, as I say to to 16 00:01:01,680 --> 00:01:05,519 Mike at the beginning of the of the chat is his name often comes 17 00:01:05,519 --> 00:01:07,800 up when I speak to, to other people. 18 00:01:08,320 --> 00:01:11,760 And I was always really intrigued to, to speak to him, 19 00:01:12,440 --> 00:01:18,200 particularly because when I was at Oxford myself, he was always 20 00:01:18,200 --> 00:01:21,120 a figure that was, that was mentioned and was influential in 21 00:01:21,120 --> 00:01:26,400 so many bits. Probably for most people, he is 22 00:01:26,720 --> 00:01:32,040 most known as the, I guess you'd call it the lead developer 23 00:01:32,040 --> 00:01:37,040 instigator of the Hydra Rolls Royce CFD code that is used 24 00:01:37,040 --> 00:01:42,400 stilted today by I would assume, thousands of engineers around 25 00:01:42,400 --> 00:01:46,400 the world to design the the jet engines that they produce. 26 00:01:46,920 --> 00:01:52,640 And, and I think what makes really interesting is his pivot 27 00:01:52,720 --> 00:01:59,840 also to them work in finance, computational maths in in as 28 00:01:59,840 --> 00:02:07,440 applied to, to, to, to finance, where he is equally made in 29 00:02:07,440 --> 00:02:08,840 prayer. And that's what's incredible for 30 00:02:08,840 --> 00:02:15,480 someone to make an impression in two quite different fields, CFD 31 00:02:15,520 --> 00:02:19,640 and and finance being quite different shows the level of the 32 00:02:19,640 --> 00:02:25,760 person, the intellect, the the sort of innovation potential to 33 00:02:25,760 --> 00:02:29,040 switch to a different field and then still make an impact. 34 00:02:29,760 --> 00:02:34,160 Now I only come from the first field, so my appreciation of the 35 00:02:34,160 --> 00:02:36,920 second in the sort of quant side of the world. 36 00:02:36,920 --> 00:02:41,200 But I've read enough to know and you only have to look at do a 37 00:02:41,200 --> 00:02:47,080 quick Google search to see that he's his work on what's called 38 00:02:47,080 --> 00:02:52,280 that multi level Monte Carlo methods has been very impactful 39 00:02:52,280 --> 00:02:57,040 in that community. And the third thing that I guess 40 00:02:57,040 --> 00:03:01,360 he is really pioneered in some ways, and one one was one of the 41 00:03:01,360 --> 00:03:04,600 early adopters for is on the high, high performance computing 42 00:03:04,600 --> 00:03:07,120 side and particularly around GPUs. 43 00:03:08,280 --> 00:03:12,800 I should state obviously for transparency, I do now work at 44 00:03:12,800 --> 00:03:17,720 that company NVIDIA, but this conversation was in no way 45 00:03:17,720 --> 00:03:22,120 arranged to to promote NVIDIA. This was organized completely 46 00:03:22,120 --> 00:03:24,760 separately. And it just happens to be that 47 00:03:24,760 --> 00:03:28,520 I, I, I don't I would there. So you'll hear some mentions of 48 00:03:28,520 --> 00:03:31,560 it, but please trust me, this is not some sort of, you know, 49 00:03:32,000 --> 00:03:35,440 product placement. And and he was actually working 50 00:03:35,440 --> 00:03:38,640 on this and I think he said that he was maybe the number 2 or 51 00:03:38,640 --> 00:03:42,280 like the second ever CUDA fellow and was looking at this in 2000 52 00:03:42,280 --> 00:03:46,200 and six 2007. You know, well before sort of 53 00:03:46,200 --> 00:03:49,280 everybody knows the name of video and GPU. 54 00:03:49,280 --> 00:03:53,480 So he, he is always, and he says in the discussion, had an 55 00:03:53,480 --> 00:03:56,880 interest in high performance computers throughout the time 56 00:03:57,120 --> 00:04:01,680 and, and still today teaches classes and, and programs and 57 00:04:01,680 --> 00:04:04,160 does everything, which is, I really love that when you see 58 00:04:04,160 --> 00:04:07,040 someone who's gone through their whole career and it's completely 59 00:04:07,040 --> 00:04:09,000 fine if you do change. And some people as they 60 00:04:09,000 --> 00:04:12,400 progress, become more senior, they get less hands on, you 61 00:04:12,400 --> 00:04:15,600 know, and then they're more about enacting a vision for what 62 00:04:15,600 --> 00:04:17,079 they want to do. And there's nothing wrong with 63 00:04:17,079 --> 00:04:19,600 that. But I always have a special 64 00:04:19,600 --> 00:04:23,000 appreciation for, for people who are, you know, one of the 65 00:04:23,000 --> 00:04:28,360 world's leading professors and they, they're still a hands on, 66 00:04:28,560 --> 00:04:30,640 you know, keyboard. So yeah. 67 00:04:32,640 --> 00:04:37,160 Mike is a professor of numerical analysis at the Maths Institute 68 00:04:37,400 --> 00:04:40,280 at the University of Oxford. As we mentioned, beautiful 69 00:04:40,280 --> 00:04:42,320 building, lovely location. I'm very jealous. 70 00:04:44,000 --> 00:04:48,880 And he was at the University of Cambridge that was rated as an 71 00:04:48,880 --> 00:04:51,000 undergraduate. And we mentioned it, he was a 72 00:04:51,000 --> 00:04:53,960 senior Wrangler, which if you look it up on Wikipedia, 73 00:04:53,960 --> 00:04:57,920 basically means the person who graduated top of class for the 74 00:04:57,920 --> 00:05:00,600 whole university in terms of maths as an undergraduate. 75 00:05:01,320 --> 00:05:04,320 And then he went to MIT, he was Kennedy Scholar, taught there, 76 00:05:04,680 --> 00:05:07,120 came back to Oxford. Now we go through all of that, 77 00:05:07,680 --> 00:05:11,200 but one of the things that's probably worth mentioning and 78 00:05:11,200 --> 00:05:15,200 congratulating him on is that actually very recently he was 79 00:05:15,200 --> 00:05:18,280 elected a fellow of the Royal Society, which is one of the 80 00:05:18,280 --> 00:05:21,120 highest honors that can be bestowed on somebody. 81 00:05:21,720 --> 00:05:25,880 So he's had an amazing career, still has an amazing career. 82 00:05:26,200 --> 00:05:30,160 And like any of these episodes, when I talk to something like 83 00:05:30,160 --> 00:05:33,440 that, there are so many things that I realized I didn't ask him 84 00:05:33,440 --> 00:05:36,920 after I finished the episode. And I don't think we would have 85 00:05:36,920 --> 00:05:39,360 time anyway to, to to go through stuff. 86 00:05:40,280 --> 00:05:43,920 I will just note that if you Google his name and go to his 87 00:05:43,920 --> 00:05:47,400 personal website, I'll put it in the chat for the YouTube side of 88 00:05:47,400 --> 00:05:49,720 things. He has a great link to lots of 89 00:05:49,720 --> 00:05:52,120 presentations courses that he's done. 90 00:05:52,840 --> 00:05:57,040 So I would definitely look at there on that to yeah, to to 91 00:05:57,240 --> 00:06:00,000 find out more. But yeah, I really hope you 92 00:06:00,000 --> 00:06:02,440 enjoy this conversation. I genuinely did. 93 00:06:02,520 --> 00:06:04,720 I hope you can see it from my face if you're watching it. 94 00:06:04,720 --> 00:06:08,240 I was learning and, you know, interested throughout the whole 95 00:06:08,480 --> 00:06:11,160 2 hours. So yeah, please sit back and 96 00:06:11,160 --> 00:06:13,720 enjoy this episode with Professor Mike Childs. 97 00:06:14,480 --> 00:06:16,440 Thank you very much. Really appreciate. 98 00:06:16,720 --> 00:06:20,440 As I said just before your name comes up a lot in people I speak 99 00:06:20,440 --> 00:06:25,480 to in the in the CFD and HPC world. 100 00:06:25,480 --> 00:06:29,520 And so I was, yeah, really wanting to speak to you to find 101 00:06:29,520 --> 00:06:32,520 out a little bit more how this, you know, connections started 102 00:06:32,520 --> 00:06:38,880 out. But maybe we could start, you 103 00:06:38,880 --> 00:06:41,720 know, I guess towards towards the beginning. 104 00:06:41,720 --> 00:06:44,480 Now you're a professor at, you know, one of the top 105 00:06:44,480 --> 00:06:46,960 universities in the in the in the world and maths department, 106 00:06:46,960 --> 00:06:48,640 but obviously with an engineering background. 107 00:06:50,520 --> 00:06:54,240 Did you always want? Was your interest in maths? 108 00:06:54,360 --> 00:06:56,440 Engineering? Were you a person with planes 109 00:06:56,440 --> 00:06:58,560 and cars? More reading? 110 00:06:58,560 --> 00:07:01,040 Physics What? What was your early days like? 111 00:07:02,680 --> 00:07:07,320 So I would say in school my interests were maths and 112 00:07:07,320 --> 00:07:14,200 physics. And so I actually, you know, 113 00:07:14,400 --> 00:07:19,160 went to Cambridge to study maths, intending after the first 114 00:07:19,160 --> 00:07:22,080 year to transfer into theoretical physics. 115 00:07:24,200 --> 00:07:28,240 But then, you know, enjoyed the maths, stayed, stayed in the 116 00:07:28,240 --> 00:07:32,080 maths. So, so as a child, that's kind 117 00:07:32,080 --> 00:07:37,280 of the direction I suppose I saw myself going in, but I also 118 00:07:37,280 --> 00:07:41,720 wasn't sure, you know, what I would want to do after my 119 00:07:41,720 --> 00:07:46,160 studies. And so in terms of CFDI mean 120 00:07:46,160 --> 00:07:52,960 really the, the pivotal thing for me was the fact that in in 121 00:07:52,960 --> 00:07:57,000 going to Cambridge in those days, you did the entrance exam 122 00:07:57,000 --> 00:08:02,160 in November, December and then you had from January to October 123 00:08:02,160 --> 00:08:06,360 to do something else. And some people have travelled 124 00:08:06,360 --> 00:08:12,200 then in my case, I went and worked at Rolls Royce for that 125 00:08:12,200 --> 00:08:16,440 period. And that, that was really to 126 00:08:16,440 --> 00:08:20,640 learn well, well, to see what engineering was like, see if 127 00:08:20,640 --> 00:08:22,880 that's something which interested me. 128 00:08:24,320 --> 00:08:29,120 So at age 18, I was an Rolls Royce undergraduate engineering 129 00:08:29,120 --> 00:08:32,400 apprentice. That was my, my job title. 130 00:08:33,919 --> 00:08:35,960 And normally they wouldn't, wouldn't have taken a 131 00:08:35,960 --> 00:08:40,000 mathematician on, you know, usually it, it was engineers, 132 00:08:40,000 --> 00:08:45,040 maybe people in materials. But I was actually third 133 00:08:45,040 --> 00:08:48,920 generation Rolls Royce. My, my mother was a programmer 134 00:08:48,920 --> 00:08:51,400 before I was born. She was a programmer with Rolls 135 00:08:51,400 --> 00:08:54,840 Royce in Derby. And my grandfather worked, 136 00:08:54,840 --> 00:08:58,840 worked for the company for more than 25 years in Glasgow. 137 00:09:00,360 --> 00:09:05,440 So back in those days, you know, the application forms asked if 138 00:09:05,440 --> 00:09:08,000 you had any family members working in the company. 139 00:09:08,200 --> 00:09:11,400 I mean, these days that, that that would be nepotism that's 140 00:09:11,680 --> 00:09:14,880 strictly forbidden. But back in those days that that 141 00:09:14,880 --> 00:09:19,960 was viewed as a positive thing. And so I think because of that, 142 00:09:19,960 --> 00:09:26,520 they, they, they took me on as an apprentice because I was a 143 00:09:26,520 --> 00:09:28,720 mathematician. They weren't exactly sure what 144 00:09:28,720 --> 00:09:31,760 to do with me. So, so I went through a lot of 145 00:09:31,760 --> 00:09:36,400 the standard training with, with the engineers then, whereas the 146 00:09:36,400 --> 00:09:39,840 engineers had to be moved around different parts of the company 147 00:09:40,280 --> 00:09:43,640 to satisfy, you know, requirements for chartered 148 00:09:43,640 --> 00:09:47,280 engineer status. You know, later on, you know, I 149 00:09:47,280 --> 00:09:50,760 had, I had more flexibility as, as to what I did. 150 00:09:50,760 --> 00:09:55,040 And so I got into various assignments which involved 151 00:09:55,040 --> 00:10:00,760 programming in in various forms. Oh wow, that's interesting. 152 00:10:00,760 --> 00:10:03,800 So this is before you did your undergraduate? 153 00:10:03,840 --> 00:10:06,840 Yes, this is the period after school. 154 00:10:07,960 --> 00:10:11,360 Before undergraduate and then each summer while I was at 155 00:10:11,360 --> 00:10:14,840 Cambridge, each summer I went back to Rolls Royce for, for 156 00:10:14,840 --> 00:10:20,960 another two months. So it was after my second year. 157 00:10:20,960 --> 00:10:27,040 So that would be 1980 that I joined what was essentially the 158 00:10:27,040 --> 00:10:31,600 CFD group, you know, only a few months after it was first 159 00:10:31,680 --> 00:10:34,880 created, you know, so it was called the theoretical sciences 160 00:10:34,880 --> 00:10:41,280 group. And that that summer I was doing 161 00:10:42,080 --> 00:10:45,160 2D grid generation using conformal mapping. 162 00:10:47,320 --> 00:10:52,120 So directly using, using my, my, my coursework in complex 163 00:10:52,120 --> 00:10:57,080 variable theory. That was actually a paper, I 164 00:10:57,080 --> 00:11:00,080 think it was written by Bob Nee at Pratt and Whitney. 165 00:11:00,640 --> 00:11:03,560 I don't know whether you recognise that, that name, but 166 00:11:04,520 --> 00:11:08,560 Bob's a really, you know, senior figure of that of that era. 167 00:11:08,560 --> 00:11:09,800 Oh. Wow. 168 00:11:10,000 --> 00:11:13,320 OK, so you did. So you're going between, but in 169 00:11:13,320 --> 00:11:18,520 the during your undergraduate, did you already therefore get a 170 00:11:18,520 --> 00:11:24,480 sense that you wanted to go down the more CFD route because you 171 00:11:24,480 --> 00:11:27,400 were doing in Rolls Royce? I imagine because you were doing 172 00:11:27,400 --> 00:11:31,960 maths you didn't touch. Well, I guess would you do fluid 173 00:11:31,960 --> 00:11:34,320 dynamics in? I guess you would do numerical 174 00:11:34,320 --> 00:11:36,480 methods, but would would fluids come up? 175 00:11:37,080 --> 00:11:40,320 So. So we did lots of theoretical 176 00:11:40,320 --> 00:11:44,680 fluid dynamics. So I can remember thinking, you 177 00:11:44,720 --> 00:11:48,480 know, on on the one hand here, here I am taking a course 178 00:11:48,720 --> 00:11:53,960 learning about invisid incompressible 2D flow over a 179 00:11:53,960 --> 00:11:56,840 cylinder. And on the other hand, here I am 180 00:11:56,840 --> 00:12:03,720 at Rolls Royce looking at these phenomenally complex, you know, 181 00:12:03,720 --> 00:12:09,960 engineering devices. You're clearly paper and pencil 182 00:12:09,960 --> 00:12:14,560 cannot take you very far. So I was really sold on the idea 183 00:12:14,560 --> 00:12:18,400 of numerical simulation at a very early age. 184 00:12:18,760 --> 00:12:20,880 You know, a very, very early stage. 185 00:12:22,000 --> 00:12:26,760 I mean, already by the time I went to university I'd done some 186 00:12:26,760 --> 00:12:30,200 amount of programming. So I got into programming pretty 187 00:12:30,200 --> 00:12:34,720 early in, in, in part through, through my mother. 188 00:12:34,720 --> 00:12:38,800 I mean she, she was IT support at University of Sterling. 189 00:12:39,920 --> 00:12:43,800 So in the holidays I actually did a little bit of programming 190 00:12:43,800 --> 00:12:48,360 on one of the big academic systems down in Manchester. 191 00:12:51,120 --> 00:12:53,800 Where was that in the big building? 192 00:12:56,200 --> 00:13:00,720 When was that in Manchester? I mean, I, I didn't go there. 193 00:13:00,720 --> 00:13:03,160 I mean, this was remote access from Sterling. 194 00:13:04,760 --> 00:13:07,000 Yeah. So this was in, in, in the 195 00:13:07,000 --> 00:13:09,800 sense, you know, the equivalent of what's now the Edinburgh 196 00:13:09,920 --> 00:13:12,720 Apparel Computing Centre. Back in those days, it was the 197 00:13:12,720 --> 00:13:18,240 Manchester centre. I don't remember what I did on 198 00:13:18,240 --> 00:13:20,200 it. Nothing very significant. 199 00:13:21,520 --> 00:13:24,280 But also while, while that Rolls Royce for those eight months 200 00:13:24,280 --> 00:13:28,920 before going to Cambridge one one day a week, they sent us 201 00:13:28,920 --> 00:13:33,360 along to what was then called Darby Tech to sort of keep our, 202 00:13:33,720 --> 00:13:37,320 our brains ticking over doing various classes, including 203 00:13:37,320 --> 00:13:41,520 programming. And so I, I, I worked in an IBM 204 00:13:41,840 --> 00:13:48,080 system there, I, I wrote a code to do project critical path 205 00:13:48,080 --> 00:13:53,360 analysis, which was great fun. So so I've always enjoyed 206 00:13:53,360 --> 00:13:54,480 programming. What? 207 00:13:54,480 --> 00:13:56,680 What languages would it be? I'm sorry if I'm asking a stupid 208 00:13:56,680 --> 00:13:57,840 question, I'm trying. To. 209 00:13:58,040 --> 00:14:03,640 Get back to like. So I think that must have been 210 00:14:03,640 --> 00:14:06,280 Fortran. It was punch cards. 211 00:14:07,360 --> 00:14:11,600 So that was my experience with, with, with punch cards that that 212 00:14:11,600 --> 00:14:16,360 time at Darby Tech, it was, it was a cast off the IBM machine 213 00:14:16,360 --> 00:14:18,680 from Rolls Royce. They, they, they donated it to 214 00:14:18,680 --> 00:14:22,160 Darby Tech. That's, that's my recollection 215 00:14:22,160 --> 00:14:26,920 anyway. So yes, I think that must have 216 00:14:26,920 --> 00:14:33,800 been Fortran in, in Cambridge, we had the whole teaching lab of 217 00:14:33,800 --> 00:14:39,160 desktop machines and that was basic, I think that we used. 218 00:14:39,680 --> 00:14:43,480 So there was a, there was a numerical projects course in the 219 00:14:43,480 --> 00:14:47,520 third year, which was very good, you know, and so that, that 220 00:14:47,520 --> 00:14:51,920 really sort of solidified my, my, my interest in, in, in 221 00:14:51,920 --> 00:15:00,920 computational methods. And then what about so I, I, I 222 00:15:00,920 --> 00:15:02,960 have to ask you one, one question. 223 00:15:02,960 --> 00:15:05,680 When I was doing a bit of research, I hadn't come across 224 00:15:05,680 --> 00:15:10,400 this term before, but it's quite an esteemed senior Wrangler. 225 00:15:10,640 --> 00:15:12,000 Am I pronouncing it? Correctly. 226 00:15:12,920 --> 00:15:14,400 I guess you've looked at the Wikipedia. 227 00:15:15,800 --> 00:15:19,320 And then I was like, then I started to look down and I read 228 00:15:19,520 --> 00:15:22,360 fantastic stories about people being paraded around. 229 00:15:22,360 --> 00:15:24,240 I don't know if that was the time. 230 00:15:24,240 --> 00:15:29,320 So this was the top undergraduate of maths, is that 231 00:15:29,320 --> 00:15:31,960 correct? Which which you you've got. 232 00:15:32,680 --> 00:15:35,680 Yeah, there there was no fanfare in my game. 233 00:15:40,000 --> 00:15:43,760 So yeah, I was just told afterwards by by my tutor, yeah. 234 00:15:44,320 --> 00:15:49,160 OK, well, it's it's still it shows, I guess your your 235 00:15:49,160 --> 00:15:55,480 ability. Were you, did you enjoy the, the 236 00:15:55,480 --> 00:15:58,000 sort of Cambridge life, the collegiate life? 237 00:15:58,040 --> 00:16:01,960 Was, was that something that you, I mean, now obviously you 238 00:16:01,960 --> 00:16:05,520 know, you're at the other place, but was that something that you 239 00:16:05,880 --> 00:16:09,240 made a, a strong impression on you and, and sort of motivated 240 00:16:09,240 --> 00:16:13,800 you later on to, to ultimately stay in or, or go back to 241 00:16:13,800 --> 00:16:18,960 academia? I mean, I, I enjoyed my time at 242 00:16:18,960 --> 00:16:23,200 Cambridge. I spent a lot of time doing 243 00:16:23,920 --> 00:16:27,760 orienteering while I was there. So orienteering was something I 244 00:16:27,760 --> 00:16:31,800 did as a child, you know, from the age of about 12, you know, 245 00:16:31,800 --> 00:16:34,280 for. Yeah, for, for those listening 246 00:16:34,280 --> 00:16:36,880 to this who don't know about orienteering, it's effectively 247 00:16:37,160 --> 00:16:41,560 cross country running, using a map to guide yourself through 248 00:16:41,560 --> 00:16:43,720 forests and over hillsides and stuff. 249 00:16:45,400 --> 00:16:50,720 So, yeah, a lot of my time at Cambridge was spent going off 250 00:16:51,280 --> 00:16:59,720 the orienteering at weekends. I would say that I was as a 251 00:16:59,720 --> 00:17:03,040 diligent student, but I wasn't particularly hard working. 252 00:17:03,040 --> 00:17:04,640 Let's let's let's put it that way. 253 00:17:04,880 --> 00:17:10,839 You know, I seem to remember my, my maths as being sort of a, a 254 00:17:10,839 --> 00:17:14,079 nine to five activity five days a week. 255 00:17:14,079 --> 00:17:15,880 And then, you know, weekends I was away. 256 00:17:17,319 --> 00:17:20,119 I didn't really start working until I went to MIT. 257 00:17:20,720 --> 00:17:25,119 OK, that, that that sort of sums up the difference between MIT 258 00:17:25,119 --> 00:17:28,680 and Cambridge also, you know, grad student life and undergrad 259 00:17:28,680 --> 00:17:30,920 life, I guess. Yeah, Yeah. 260 00:17:30,920 --> 00:17:34,960 So you said the beginning that you'd, you weren't sure. 261 00:17:34,960 --> 00:17:38,040 You, you were debating around theoretical physics, maths I 262 00:17:38,040 --> 00:17:40,560 presume. Then as you went towards the end 263 00:17:40,560 --> 00:17:45,400 of your undergrad and particularly with the summers in 264 00:17:45,400 --> 00:17:49,360 Rolls Royce, you'd, you'd put aside the theoretical physics 265 00:17:49,480 --> 00:17:53,360 and you were more moving towards, I guess the engineering 266 00:17:53,360 --> 00:17:55,760 or applied mathematics. Would that be fair? 267 00:17:56,240 --> 00:18:00,120 I mean, I think by the time I finished that initial 8 months 268 00:18:00,120 --> 00:18:08,200 at Rolls Royce, I think probably the physics ideas had to a large 269 00:18:08,200 --> 00:18:13,160 extent dropped out. Although I think equally I knew 270 00:18:13,160 --> 00:18:15,200 I didn't want to work in industry. 271 00:18:15,760 --> 00:18:19,320 So yeah. So maybe things were still 272 00:18:19,320 --> 00:18:28,400 somewhat open. Yes, I I do remember, I'm trying 273 00:18:28,400 --> 00:18:31,680 to think whether it would be end of my second year or early in my 274 00:18:31,680 --> 00:18:37,000 third year. There was a talk in college by 275 00:18:38,600 --> 00:18:43,120 an academic, A researcher just talking about the joy of being 276 00:18:43,120 --> 00:18:47,640 an academic researcher and that that did strike me. 277 00:18:48,200 --> 00:18:54,440 So, so that's one of those moments that I think, you know, 278 00:18:54,480 --> 00:18:59,960 confirmed me in going as an academic path or, or at least 279 00:18:59,960 --> 00:19:02,800 taking it further and doing, doing a PhD. 280 00:19:06,720 --> 00:19:13,560 But you know, the move to MIT, my, my tutor told me about the, 281 00:19:13,560 --> 00:19:18,240 the Kennedy scholarship scheme and encouraged me to apply, you 282 00:19:18,240 --> 00:19:22,000 know, and so, you know, initially I went to MIT on, on 283 00:19:22,000 --> 00:19:26,800 this one year scholarship thinking it would be a chance to 284 00:19:26,800 --> 00:19:30,680 see the world. And then I would think about 285 00:19:31,040 --> 00:19:38,200 what to do next. Having gone out to MIT, my, my 286 00:19:38,200 --> 00:19:43,080 supervisor there found funding for my second year to, to finish 287 00:19:43,080 --> 00:19:48,600 up the masters. And in doing that, I also came 288 00:19:48,600 --> 00:19:53,200 up with a, what turned out to be a good idea for a PhD project. 289 00:19:53,640 --> 00:19:57,640 And then, then he got more, more research funding for, for me 290 00:19:59,080 --> 00:20:03,640 actually from US Air Force to, to carry on and do do the PhD. 291 00:20:04,200 --> 00:20:07,480 So it wasn't. It wasn't the master plan by any 292 00:20:07,480 --> 00:20:13,040 means, but yeah, yeah, I went out initially for one year and 293 00:20:13,040 --> 00:20:17,360 ended up staying for 11. So what was the Kennedy 294 00:20:17,360 --> 00:20:18,920 scholarship like? Because I spoke to somebody 295 00:20:18,920 --> 00:20:21,880 else, Anthony. Oh, sorry, I'm confusing. 296 00:20:24,120 --> 00:20:29,240 Oh, got a brain fog. Now the have to rise. 297 00:20:31,000 --> 00:20:36,720 Now I'm just interested because I wonder if you'd actually come 298 00:20:36,960 --> 00:20:39,480 across. Probably you haven't. 299 00:20:40,640 --> 00:20:44,440 My my memory for names is terrible, but there wasn't 300 00:20:44,440 --> 00:20:46,880 Anthony about my time. Yeah, that's why I'm. 301 00:20:47,600 --> 00:20:51,280 Just. Wondering if you've in In 302 00:20:51,280 --> 00:20:54,640 general, most of the Kennedy scholars were at Harvard. 303 00:20:54,640 --> 00:21:00,040 There were very few at MIT and, and, and these days it's really 304 00:21:00,040 --> 00:21:03,920 quite the exception to have anybody at MIT, which is a bit 305 00:21:03,920 --> 00:21:06,640 of a shame I think. Yeah, Tony Pannell. 306 00:21:07,320 --> 00:21:13,200 Did you ever come across Tony? Yeah, that that that name does 307 00:21:13,200 --> 00:21:14,560 sound familiar. So what? 308 00:21:14,600 --> 00:21:15,880 What? What's he doing now? 309 00:21:16,280 --> 00:21:19,800 Yeah, So Tony, who I actually also interviewed, he's a really 310 00:21:19,920 --> 00:21:23,480 great guy. So he he won the Kennedy 311 00:21:23,480 --> 00:21:30,440 scholarship, went to MIT and I think it was in 80, mid 80s, so 312 00:21:30,440 --> 00:21:36,280 similar time. And then he went to he then went 313 00:21:36,280 --> 00:21:37,960 to work. He created his own company, but 314 00:21:37,960 --> 00:21:41,240 he ultimately ended up in Formula One running the what is 315 00:21:41,240 --> 00:21:43,280 now the Red Bull team. But he's now a professor at 316 00:21:43,280 --> 00:21:46,600 Cambridge. Does all the aerodynamics and 317 00:21:46,600 --> 00:21:51,280 CFDII wonder whether? Yeah, so, yeah. 318 00:21:51,280 --> 00:21:54,680 So I was a Kennedy scholar 81 to 82. 319 00:21:58,000 --> 00:22:01,040 Your and the name does sound familiar. 320 00:22:01,320 --> 00:22:04,960 Basically, we we didn't really hang out with each other. 321 00:22:05,520 --> 00:22:08,240 I didn't hang out with the other Kennedy scholars. 322 00:22:10,400 --> 00:22:15,880 So yes, yes. I didn't really have have have 323 00:22:15,880 --> 00:22:18,960 those connections, but it was a part time experience. 324 00:22:18,960 --> 00:22:22,560 Oh, oh, yes, yeah, yeah. So, so, you know, it was Kennedy 325 00:22:22,560 --> 00:22:24,880 scholarship that took me over to MIT. 326 00:22:25,480 --> 00:22:28,400 And as it happened, and this this was really sort of 327 00:22:28,400 --> 00:22:34,600 coincidence, my supervisor at MIT had Rolls Royce funding. 328 00:22:35,920 --> 00:22:41,400 This was in the days when Rolls Royce was starting to sell 329 00:22:41,400 --> 00:22:46,480 engines to the US Marine Corps for the Harrier, and they wanted 330 00:22:46,480 --> 00:22:49,840 to be viewed as more of an international company. 331 00:22:51,040 --> 00:22:56,040 And so as part of that, I think almost out of their marketing 332 00:22:56,040 --> 00:22:59,200 budget, maybe there was a whole pile of research funding to be 333 00:22:59,200 --> 00:23:04,960 spent to MIT. And and so some of this was was 334 00:23:04,960 --> 00:23:09,040 going to my supervisor guy guy by the name of Tilt Tompkins. 335 00:23:12,240 --> 00:23:16,760 So although my going to MIT and, you know, landing up with Tilt 336 00:23:16,760 --> 00:23:19,800 as a supervisor was completely independent of Rolls Royce, 337 00:23:20,160 --> 00:23:23,520 there was still that sort of accidental background 338 00:23:23,520 --> 00:23:32,520 connection. So then my, my, my, my graduate 339 00:23:32,520 --> 00:23:36,040 history at, at MIT is a bit curious. 340 00:23:36,040 --> 00:23:42,120 So I did did the masters in 18 months, which is a bit faster 341 00:23:42,120 --> 00:23:44,880 than normal, but not exceptional. 342 00:23:45,480 --> 00:23:50,800 I then did did my PhD in 2 1/2 years, which is highly unusual 343 00:23:50,800 --> 00:23:55,520 for MIT. So, So what happened there was I 344 00:23:55,520 --> 00:24:01,400 was basically two years into my PhD and you know, it had gone 345 00:24:01,400 --> 00:24:04,040 very well. I mean, this was, you know, the 346 00:24:04,040 --> 00:24:07,720 project that me and Mark Drella did, you know, you know, the ICS 347 00:24:07,720 --> 00:24:11,120 code, you know, 2D airfoil design code. 348 00:24:13,200 --> 00:24:17,320 So the project was going ahead very well, but I got called into 349 00:24:18,080 --> 00:24:24,520 the head of departments office about two years into my PhD and 350 00:24:24,520 --> 00:24:28,680 told that my supervisor hadn't got tenure and would be leaving 351 00:24:28,680 --> 00:24:31,400 in six months time and would I like his job. 352 00:24:34,720 --> 00:24:39,040 Wow. So so I then had to finish up 353 00:24:39,040 --> 00:24:42,520 really quickly that that I wasn't. 354 00:24:42,520 --> 00:24:43,400 Expecting that. Wow. 355 00:24:43,560 --> 00:24:46,000 OK, so. Yeah, yes, that, that, that, 356 00:24:46,000 --> 00:24:48,920 that last six months was, yes, exhausting. 357 00:24:50,200 --> 00:24:53,840 So then, so your your pH. What was the? 358 00:24:53,920 --> 00:24:57,160 What was the end title of your thesis then for the? 359 00:24:57,560 --> 00:25:02,480 Something like 2 dimensional transonic aerodynamic design 360 00:25:02,480 --> 00:25:06,640 method. So did you work with or be 361 00:25:06,640 --> 00:25:10,840 inspired, I guess by, by Professor Jameson? 362 00:25:11,200 --> 00:25:14,520 Was there, was there any sort of link into? 363 00:25:15,000 --> 00:25:17,400 I just, I always find it interesting when there's sort of 364 00:25:17,400 --> 00:25:20,840 Brits going over to the US. Yeah, yeah, they go and work on 365 00:25:20,840 --> 00:25:22,160 things. But it was independent. 366 00:25:22,160 --> 00:25:24,120 There was no he didn't have a. Connection. 367 00:25:24,120 --> 00:25:30,080 No, there was no connection at all there, although he, he, he 368 00:25:30,080 --> 00:25:35,760 was aware of me. He was aware of my master's 369 00:25:35,760 --> 00:25:42,720 thesis and actually told, told tilt type, you know, something 370 00:25:42,720 --> 00:25:45,240 along the lines of, you know, they should have given me a PhD 371 00:25:45,240 --> 00:25:52,360 for it, which it was flattering. But but no, my, my, my master's 372 00:25:52,360 --> 00:25:59,560 thesis was using some mathematics WKB analysis to 373 00:25:59,560 --> 00:26:05,800 understand some numerical wave propagation on grids. 374 00:26:06,360 --> 00:26:08,840 And what it shows you is that if you have something like the 375 00:26:08,840 --> 00:26:14,080 convection equation, if you have a poorly resolved wave, it can 376 00:26:14,080 --> 00:26:17,680 actually travel in the wrong direction and you can get these 377 00:26:17,680 --> 00:26:20,640 weird wave track wave trapping phenomenon. 378 00:26:20,960 --> 00:26:23,720 Anyway, it, it, it, it, it was something that intrigued 379 00:26:23,720 --> 00:26:28,840 Anthony. And so, so he was aware of me 380 00:26:28,840 --> 00:26:34,600 already at at that point. I mean, I was obviously aware of 381 00:26:34,760 --> 00:26:42,480 him, but the, you know, the, the stream tube idea that I had, 382 00:26:42,480 --> 00:26:45,960 which was the basis of the ICS code, that was completely 383 00:26:45,960 --> 00:26:53,520 different to anything, you know, in, in, in CFD at that time, you 384 00:26:53,560 --> 00:26:58,760 know, worked beautifully in 2D. It had had no natural 3D Ext, 385 00:26:59,600 --> 00:27:04,120 but, but for, for 2D wing design it, it, it was really ideal. 386 00:27:04,440 --> 00:27:07,640 And so Mark, Mark Drella and I teamed up on that. 387 00:27:08,040 --> 00:27:12,920 And in fact, for a while we were going to do a joint thesis. 388 00:27:13,120 --> 00:27:16,640 We're actually going to write it up as a single document between 389 00:27:16,640 --> 00:27:19,960 the two of us. And our thesis committee was 390 00:27:20,320 --> 00:27:23,440 perfectly happy with this. And since our thesis committee 391 00:27:23,440 --> 00:27:27,320 included the head of department, I assumed that this was all 392 00:27:27,320 --> 00:27:29,440 fine. And then later on in the 393 00:27:29,440 --> 00:27:35,360 process, the central university somehow, you know, learnt about 394 00:27:35,360 --> 00:27:38,600 this and said there is no precedent for this, there shall 395 00:27:38,600 --> 00:27:41,160 be no precedent. You know, you're not allowed. 396 00:27:42,520 --> 00:27:46,080 And so very late in the process, basically, you know, Mark and I 397 00:27:46,080 --> 00:27:50,480 had to sort of carve our work up into two separate pieces so that 398 00:27:50,480 --> 00:27:52,440 we could write up two separate documents. 399 00:27:53,160 --> 00:27:57,000 You know, so in, in, in the end, I, I finished up early because I 400 00:27:57,000 --> 00:27:58,920 had to take over Tilt's position. 401 00:27:59,160 --> 00:28:03,360 Mark, Mark finished up a year later then, you know, he did, he 402 00:28:03,360 --> 00:28:06,280 got an academic position as well, you know, so we were both 403 00:28:06,280 --> 00:28:08,040 hired. I mean, there was never any 404 00:28:08,080 --> 00:28:11,160 question of one of us freeloading off the other one. 405 00:28:11,160 --> 00:28:14,680 I mean, our, our thesis committee were perfectly content 406 00:28:14,680 --> 00:28:18,480 on that point. So yes, so it's a curious 407 00:28:18,480 --> 00:28:25,840 situation, but it meant that I, I took over Tilt's office, PA 408 00:28:26,560 --> 00:28:31,240 software engineer, almost all of his students and all of his 409 00:28:31,240 --> 00:28:34,360 research contracts, including the Rolls Royce research 410 00:28:34,360 --> 00:28:37,280 contract. So, so at that point I was then 411 00:28:37,280 --> 00:28:38,960 back into the Rolls Royce family. 412 00:28:39,920 --> 00:28:42,680 And how old were you then? You couldn't have been that old. 413 00:28:42,680 --> 00:28:46,640 25. OK, that's quite, that's quite 414 00:28:46,640 --> 00:28:51,240 young then. In in the US system and 415 00:28:51,240 --> 00:28:57,800 particularly at MIT, they they do hire a lot of people straight 416 00:28:57,800 --> 00:29:02,880 from PhD into assistant professor positions because 417 00:29:03,160 --> 00:29:07,040 given the tenure system if they decide that they made a mistake, 418 00:29:07,040 --> 00:29:09,080 they just flush you out after seven years. 419 00:29:12,840 --> 00:29:17,200 Whereas whereas in, in the British system, you know, we, 420 00:29:17,200 --> 00:29:21,480 we, we, we like to see people get a good bit of experience 421 00:29:21,480 --> 00:29:24,840 under their belt, you know, before we'll, we'll, we'll hire 422 00:29:24,840 --> 00:29:26,600 them in, in Oxford. Yeah. 423 00:29:26,840 --> 00:29:30,440 Very occasionally we'll we'll take people straight from PhD, 424 00:29:30,440 --> 00:29:34,400 but it's very, very rare. Wow. 425 00:29:34,760 --> 00:29:38,440 So you're 25, you're an assistant professor at MIT, 426 00:29:38,440 --> 00:29:43,600 you've got APA, all that, and and now you're taking over the 427 00:29:43,600 --> 00:29:44,760 Rolls Royce. So what what? 428 00:29:45,000 --> 00:29:47,640 How do things progress from from there? 429 00:29:51,120 --> 00:29:56,800 So looking back, actually the, the, the first thing is it took 430 00:29:56,800 --> 00:30:00,720 me about six months, I think, to recover from burnout from having 431 00:30:00,720 --> 00:30:08,240 finished up the PhD so quickly. But, you know, there were, there 432 00:30:08,480 --> 00:30:11,720 were a certain number of plans that were already in place that 433 00:30:11,760 --> 00:30:14,120 I sort of carried on supervising students. 434 00:30:14,120 --> 00:30:17,440 But I guess during that first six months, I was thinking 435 00:30:17,440 --> 00:30:21,000 about, you know, what, what was the first new thing I wanted to, 436 00:30:21,160 --> 00:30:24,160 to, to do with Rolls Royce funding. 437 00:30:24,640 --> 00:30:30,080 And you know, I want, I wanted to do something different. 438 00:30:30,720 --> 00:30:34,400 I, I can't remember how much I talked to them to understand 439 00:30:34,400 --> 00:30:37,760 what their concerns were at the time. 440 00:30:37,760 --> 00:30:39,400 So that what, what, what they needed. 441 00:30:42,120 --> 00:30:48,520 But what I decided to do was to develop a 2D unsteady CFD code. 442 00:30:49,960 --> 00:30:57,480 I think that they had had some engineering challenges in I 443 00:30:57,480 --> 00:31:01,520 think it was a military engine where in military engines 444 00:31:01,520 --> 00:31:06,040 there's a smaller gap between the stators and rotors. 445 00:31:06,320 --> 00:31:09,080 So they're they're they're more closely coupled stages. 446 00:31:09,880 --> 00:31:13,480 And as that as such, that means you get a larger level of 447 00:31:13,480 --> 00:31:18,040 unsteady forcing on on on the blades. 448 00:31:18,320 --> 00:31:21,880 And I think they, that there had possibly been some engineering 449 00:31:23,040 --> 00:31:27,040 project where they had major difficulties with that and they 450 00:31:27,480 --> 00:31:31,760 really needed tools to analyse that. 451 00:31:34,720 --> 00:31:37,800 I mean, this, this was still, you know, fairly early days for, 452 00:31:37,800 --> 00:31:42,480 for, for CFD, you know, so, so I think at that point, you know, 453 00:31:42,480 --> 00:31:48,680 people like Bill Dawes and John Denton had developed steady 2D 454 00:31:48,680 --> 00:31:55,720 CFD codes, possibly even 3D, but but not anything unsteady. 455 00:31:56,160 --> 00:32:02,000 So, you know, so I, I, you know, my first code for Rolls Royce 456 00:32:02,000 --> 00:32:07,080 was one called Unsflow, which was 2D unsteady. 457 00:32:07,400 --> 00:32:11,520 Initially it was wake rotor interaction. 458 00:32:11,800 --> 00:32:14,760 So you were passing in the wakes through upstream boundary 459 00:32:14,760 --> 00:32:22,000 conditions and then going into doing stator rotor interaction. 460 00:32:22,240 --> 00:32:25,080 So you've actually got the moving blade rows, you know, 461 00:32:25,160 --> 00:32:30,240 moving relative to each other that that was initially 462 00:32:30,240 --> 00:32:33,800 envisited. I later made it viscous. 463 00:32:35,520 --> 00:32:38,160 I can't remember now what I did for a turbulence model. 464 00:32:38,160 --> 00:32:41,960 It's probably an algebraic turbulence model in those in 465 00:32:41,960 --> 00:32:44,920 those days. Mid 80s, Yeah, yeah, I guess, 466 00:32:45,160 --> 00:32:50,480 yeah. So the the the sort of unique 467 00:32:51,280 --> 00:32:56,480 thing about onsflow was this thing called the time inclined 468 00:32:57,680 --> 00:33:03,400 plane. So one of the difficulties in in 469 00:33:03,400 --> 00:33:10,680 doing unsteady analysis in turbo machinery is the number of rotor 470 00:33:10,680 --> 00:33:14,160 blades is different to the number of stator blades. 471 00:33:15,520 --> 00:33:19,320 So you want to do a simulation that just has one blade passage, 472 00:33:20,480 --> 00:33:23,760 but if you do it the natural way, that doesn't work. 473 00:33:23,760 --> 00:33:27,080 You know, you don't have the right periodicity to, to to do 474 00:33:27,080 --> 00:33:30,720 that. So the time inclined plane 475 00:33:33,000 --> 00:33:38,880 involved, you know, usually you know, when you're at time level 476 00:33:38,880 --> 00:33:42,600 N, you know, all the grid points are at the same physical time. 477 00:33:43,160 --> 00:33:47,000 The time inclined plane, you know that, that, that that time 478 00:33:47,000 --> 00:33:52,480 was inclined. So that, and you could incline 479 00:33:52,480 --> 00:33:57,080 it in such a way that you then set up the right periodicity 480 00:33:57,080 --> 00:34:00,960 condition to cope with this arbitrary blade count. 481 00:34:02,320 --> 00:34:07,800 So yeah, that that, that that was kind of the the unique 482 00:34:07,800 --> 00:34:09,679 point. And then, you know, there's also 483 00:34:09,679 --> 00:34:13,080 some maths I did on non reflecting boundary conditions 484 00:34:13,440 --> 00:34:15,920 that when you're doing these unsteady interactions with the 485 00:34:15,920 --> 00:34:21,400 boundaries very close, you want the outgoing waves to go out to 486 00:34:21,400 --> 00:34:24,000 not be artificially reflected from the boundary. 487 00:34:24,600 --> 00:34:27,520 And so there's a whole piece of research on on, on that. 488 00:34:27,920 --> 00:34:30,639 Steph, you take for granted now in a commercial solver. 489 00:34:31,679 --> 00:34:35,840 Yeah, Yeah. So I mean really I was one of 490 00:34:35,840 --> 00:34:40,480 the first people doing that in, in the context of compressible 491 00:34:40,480 --> 00:34:45,760 flow CFD yes. I mean these, these these days 492 00:34:46,320 --> 00:34:49,480 you, you've got things like absorbing boundary methods, 493 00:34:49,480 --> 00:34:53,560 which is probably what you would use if you're doing far field 494 00:34:53,760 --> 00:34:57,640 acoustics and electromagnetics. But that actually wouldn't work 495 00:34:57,640 --> 00:35:03,880 well in in this context of closely coupled stages with very 496 00:35:03,880 --> 00:35:07,240 close in boundaries. And were you always hands on? 497 00:35:07,240 --> 00:35:10,480 Were you, you know, at that time, were you always that sort 498 00:35:10,480 --> 00:35:13,040 of person who was programming it yourself? 499 00:35:13,080 --> 00:35:18,040 You had students, but you were still very much yes and and sort 500 00:35:18,040 --> 00:35:21,720 of a hands on programmer. Yes, that was hands on. 501 00:35:21,720 --> 00:35:27,800 I mean generally the students were writing their own other 502 00:35:27,800 --> 00:35:35,000 codes. So, so with Ansflow there was 503 00:35:35,000 --> 00:35:36,880 this software engineer Bob Haynes. 504 00:35:36,960 --> 00:35:39,960 So I don't know whether you recognise the name Bob Haynes. 505 00:35:40,320 --> 00:35:44,440 He, he he was responsible for developing our visualisation 506 00:35:44,440 --> 00:35:47,360 software. So visual two, Visual three. 507 00:35:48,040 --> 00:35:50,400 Yeah. Name names from the past. 508 00:35:50,920 --> 00:35:54,360 Bob's Bob's still at, at MIT, I think he still hasn't retired. 509 00:35:55,400 --> 00:35:58,960 You know, he must be about 10 years older than me, something 510 00:35:58,960 --> 00:36:04,200 like that. So, yeah. 511 00:36:04,200 --> 00:36:09,160 So I did most of the onsflow development, but Bob will have 512 00:36:09,160 --> 00:36:14,840 helped me with, with, with bits of that as well as doing, you 513 00:36:14,880 --> 00:36:17,560 know, he did, he certainly did all, all of the visualisation 514 00:36:17,720 --> 00:36:20,600 because if you're doing a, you know, 2D calculation, you know, 515 00:36:20,600 --> 00:36:23,640 you want to have some nice, nice visualisation. 516 00:36:24,040 --> 00:36:25,840 It's true then, true now, isn't it? 517 00:36:26,280 --> 00:36:31,280 So this was, this was in, in the early days of Silicon Graphics 518 00:36:31,280 --> 00:36:36,280 and there was a company called Stellar which was based just 519 00:36:36,280 --> 00:36:40,040 outside, well, in the Boston suburb. 520 00:36:41,400 --> 00:36:44,320 And I, I sort of got involved with, with them. 521 00:36:44,320 --> 00:36:48,360 And so we, we, we had a couple of their, their machines, 522 00:36:48,360 --> 00:36:55,840 beautiful machines, you know, multiple courses I recall as 523 00:36:55,840 --> 00:36:58,480 well, which was unusual. I mean, it was a very early days 524 00:36:58,480 --> 00:37:04,240 of parallel computing. And back then were you, you 525 00:37:04,240 --> 00:37:07,920 know, as much as you were developing the code for accuracy 526 00:37:07,920 --> 00:37:12,040 and the physics side, did you, did you still have a strong 527 00:37:12,040 --> 00:37:14,680 interest in the sort of high performance computing side? 528 00:37:14,680 --> 00:37:18,160 Were you were you always excited to try out different machines or 529 00:37:18,160 --> 00:37:21,200 have access to the machines or did that come later? 530 00:37:22,040 --> 00:37:26,400 No, I was always interested. So I mean, even even while I was 531 00:37:26,400 --> 00:37:30,560 doing, you know, masters and PhDs. 532 00:37:30,560 --> 00:37:34,440 So I think, oh, I'm trying to remember. 533 00:37:34,440 --> 00:37:41,360 So it was so probably soon after my master's, I spent the summer 534 00:37:41,360 --> 00:37:47,360 at NASA Langley at at at ICASE and down there I think I was 535 00:37:47,360 --> 00:37:55,960 doing programming on oh gosh, what, what what would it be? 536 00:38:00,320 --> 00:38:02,880 My mind's gone blank. There was there was Crane and 537 00:38:02,880 --> 00:38:04,200 then there was the other company. 538 00:38:05,880 --> 00:38:12,840 What was the other one called? So it was, it was before, you 539 00:38:12,840 --> 00:38:21,280 know, the ETA 10. Oh, Cyber 205. 540 00:38:21,640 --> 00:38:24,040 Yeah, that sounds, sounds maybe right. 541 00:38:25,400 --> 00:38:30,840 So I guess coming out to CDC maybe so. 542 00:38:31,080 --> 00:38:36,160 So I I had my first experience of supercomputing at at I case. 543 00:38:36,160 --> 00:38:40,560 I don't think I did a lot there. Who was there at that time? 544 00:38:40,560 --> 00:38:43,040 Because I've heard other people mention about this I case they 545 00:38:43,040 --> 00:38:46,960 don't do it anymore, I don't think, but it was this wasn't 546 00:38:47,000 --> 00:38:51,320 it. So the person who led it, I 547 00:38:51,320 --> 00:38:55,640 think when I was there was Milt Rose, but I think he may have 548 00:38:55,640 --> 00:38:58,120 retired not long after I was there. 549 00:38:58,480 --> 00:39:06,680 And then for many years it was led by somebody whose surname is 550 00:39:06,680 --> 00:39:08,200 Hosseini. I'm trying to remember what his 551 00:39:08,200 --> 00:39:11,200 first name is. Possibly Youssef. 552 00:39:12,280 --> 00:39:18,040 No, I'm not sure. Yes, yes, I case. 553 00:39:18,160 --> 00:39:23,000 I mean, you know, there are lots of academics there. 554 00:39:24,760 --> 00:39:28,600 Who else do I remember? Eli Turkel. 555 00:39:28,640 --> 00:39:32,800 I remember he was, he was there, I think, you know, the summer 556 00:39:32,800 --> 00:39:38,440 that I was there. And then, you know, we sometimes 557 00:39:38,440 --> 00:39:44,520 saw the people in, in the CFD group there and oh, there was a, 558 00:39:45,400 --> 00:39:51,320 there was a great, great person who headed up the CFD group. 559 00:39:51,320 --> 00:39:56,960 Oh, gosh, my, my, my memory today is, is is poor. 560 00:39:56,960 --> 00:40:01,280 Yeah. So, yeah. 561 00:40:01,280 --> 00:40:04,560 So, yeah. So I had, I had experience with 562 00:40:04,560 --> 00:40:14,160 this, I think it was a cyber 205 at at Langley at MIT. 563 00:40:14,240 --> 00:40:20,080 I did a little bit of work on on the Thinking Machines CM5, you 564 00:40:20,160 --> 00:40:25,280 know. That I remember at the time, you 565 00:40:25,280 --> 00:40:27,560 know, there was a lot of discussion about, you know, this 566 00:40:27,560 --> 00:40:31,080 is the future of massively parallel computing, you know so 567 00:40:31,080 --> 00:40:34,560 I think it was 64,000 processors, but each of the 568 00:40:34,560 --> 00:40:37,240 processors was incredibly elementary. 569 00:40:39,400 --> 00:40:43,760 And then it got blown away by risk computing, you know, and, 570 00:40:44,080 --> 00:40:48,120 and people putting together PC clusters, you know, so I think 571 00:40:48,120 --> 00:40:54,320 DARPA basically bankrolled it for maybe five years, eight 572 00:40:54,320 --> 00:40:57,040 years. But then, yeah, it wasn't 573 00:40:57,040 --> 00:41:01,560 capable of of sustaining itself. And then and then, you know, 574 00:41:01,560 --> 00:41:04,680 Cray, Cray really got established, you know, And so 575 00:41:05,000 --> 00:41:11,280 Cray, Cray was the winner. Yeah. 576 00:41:11,280 --> 00:41:15,240 I so Rolls Royce at one stage had the Cray. 577 00:41:15,520 --> 00:41:20,160 So I don't remember the time scales. 578 00:41:20,160 --> 00:41:25,320 So I don't remember whether Onslow was ever run on the Cray 579 00:41:25,320 --> 00:41:31,800 or not. I think it's possible it was. 580 00:41:35,760 --> 00:41:38,560 And were you travelling back when you were MIT working for 581 00:41:38,560 --> 00:41:42,200 Rolls Royce, Did you come back to the UK, to Derby to sort of 582 00:41:42,840 --> 00:41:47,000 have meetings with it or was it slightly sort of separated? 583 00:41:48,520 --> 00:41:53,440 I mean, I probably came back twice a year at most. 584 00:41:54,480 --> 00:41:59,080 So. And, you know, to some extent I 585 00:41:59,080 --> 00:42:02,040 would be coming home to come home and see people. 586 00:42:02,760 --> 00:42:05,000 And then while I was here, I would visit Rolls Royce. 587 00:42:05,000 --> 00:42:08,720 And to some extent, you know, a trip might, might be motivated, 588 00:42:09,120 --> 00:42:11,600 you know, more more primarily because of Rolls Royce, you 589 00:42:11,600 --> 00:42:16,000 know. So, yeah, I guess probably twice 590 00:42:16,000 --> 00:42:19,640 a year was the norm in those days. 591 00:42:21,280 --> 00:42:24,040 So how did things progress then at the MIT? 592 00:42:24,040 --> 00:42:28,120 So you were working on the Rolls Royce projects that, Yeah. 593 00:42:28,120 --> 00:42:30,560 How did, how did, how did things evolve during that time? 594 00:42:32,200 --> 00:42:37,320 I mean, as well as working on the Rolls Royce projects, I had 595 00:42:37,320 --> 00:42:41,640 some amount of funding from US sources. 596 00:42:41,960 --> 00:42:46,920 Not a lot, but things I think. Office of Naval Research Air 597 00:42:46,920 --> 00:42:53,400 Force Oh, we, we, we bought the stellar machines with a grant 598 00:42:53,400 --> 00:42:56,720 from DARPA. So again, DARPA was really 599 00:42:56,720 --> 00:43:01,880 active in funding new technologies to see whether, you 600 00:43:01,880 --> 00:43:03,520 know, these really were useful or not. 601 00:43:04,880 --> 00:43:10,440 That that is probably the worst proposal I have ever written. 602 00:43:12,520 --> 00:43:16,400 But but, but my excuse is by the time I submitted the proposal, 603 00:43:16,720 --> 00:43:21,760 my temperature was 102 or 103. I was going down with glandular 604 00:43:21,760 --> 00:43:24,120 fever. Oh, I've had that. 605 00:43:24,320 --> 00:43:25,960 That's bad glandular fever, isn't it? 606 00:43:26,000 --> 00:43:30,960 Yeah. So in in, in the way I was lucky 607 00:43:30,960 --> 00:43:34,440 mine was sufficiently bad that I had to be admitted into 608 00:43:34,440 --> 00:43:37,400 hospital. And so they then pumped, pumped 609 00:43:37,400 --> 00:43:41,680 me full of, you know, various antibiotics and stuff. 610 00:43:41,680 --> 00:43:47,520 And so I recovered well. I was ill for like 6 weeks or 611 00:43:47,560 --> 00:43:50,080 something, it was horrible. Yeah, I was. 612 00:43:50,880 --> 00:43:55,000 I was probably in hospital for a week or two and then I was sent 613 00:43:55,000 --> 00:44:00,400 home and told to stay at home and and recuperate for like 2 614 00:44:00,400 --> 00:44:06,280 months or something. And and So what I did was to 615 00:44:06,280 --> 00:44:13,520 write up a 60 page document documenting all of onsflow so 616 00:44:13,640 --> 00:44:19,240 that that that was the best documented code I ever wrote it. 617 00:44:19,720 --> 00:44:25,360 It's always this, this rule that people hate writing detailed 618 00:44:25,360 --> 00:44:28,880 documentation and, you know, just just laying out all the 619 00:44:28,880 --> 00:44:33,480 details of the numerics, all of the things that your future 620 00:44:33,760 --> 00:44:35,840 people modifying the code need to do. 621 00:44:35,960 --> 00:44:39,480 You know, it's very hard getting getting students to do that. 622 00:44:40,600 --> 00:44:47,640 So anyway, I yeah, yes, I'm trying to trying to remember 623 00:44:47,640 --> 00:44:50,360 exactly when, when all these different things happened. 624 00:44:50,480 --> 00:44:52,440 Yeah. So. 625 00:44:54,640 --> 00:44:58,400 So you're progressing at MIT, your assistant professor. 626 00:44:58,960 --> 00:45:02,800 Did you feel that you would always stay there or what 627 00:45:02,800 --> 00:45:06,440 started to get into your head about coming back? 628 00:45:07,240 --> 00:45:14,280 To so I, I, I had a good, good job. 629 00:45:14,280 --> 00:45:16,160 Obviously they're, they're at MIT. 630 00:45:16,160 --> 00:45:19,440 I was part of the gas turbine lab and there were wonderful 631 00:45:19,440 --> 00:45:23,320 experimentalists there. And so, you know, we, we, we did 632 00:45:23,320 --> 00:45:27,320 lots of good works of comparing numerics with experiment. 633 00:45:27,320 --> 00:45:32,400 And there's one paper we, we, we have an unsteady heat transfer 634 00:45:32,680 --> 00:45:37,480 where in the sense neither the experimentalists nor me had 635 00:45:37,520 --> 00:45:40,960 great faith in our own research and yet the results matched 636 00:45:41,280 --> 00:45:43,760 wonderfully. We were delighted. 637 00:45:46,160 --> 00:45:51,840 I, I did feel to some extent, not quite a fish out the water, 638 00:45:51,840 --> 00:45:54,560 but really I am at heart a mathematician. 639 00:45:55,040 --> 00:45:58,400 You know, you, you asked me at, at the beginning, you know, was 640 00:45:58,400 --> 00:46:01,720 I the kind of kid that tinkered with devices? 641 00:46:01,960 --> 00:46:06,880 No, I wasn't, you know, so I, I, I really wasn't by nature an 642 00:46:06,880 --> 00:46:11,880 engineer in that sense, by by nature, I'm an applied 643 00:46:11,880 --> 00:46:16,720 mathematician who enjoys mathematics and enjoys seeing it 644 00:46:16,720 --> 00:46:21,160 being useful in the real world. I'm not fundamentally at heart 645 00:46:21,160 --> 00:46:26,240 an engineer. And so in that sense, I did feel 646 00:46:26,240 --> 00:46:29,560 a little bit constrained being in the aeronautics and 647 00:46:29,560 --> 00:46:33,520 astronautics department, which is, you know, where I was at 648 00:46:33,520 --> 00:46:38,480 MIT. And that that I was sort of most 649 00:46:38,480 --> 00:46:41,240 aware of when doing things like the non reflecting boundary 650 00:46:41,240 --> 00:46:45,280 condition theory, because that was very much maths theory, but 651 00:46:45,440 --> 00:46:50,920 publishing it in in engineering CFD, So journals, you know, AI 652 00:46:50,920 --> 00:47:00,720 AA journal. So that yeah, I think also, you 653 00:47:00,720 --> 00:47:02,880 know, getting married and thinking about where, where you 654 00:47:02,880 --> 00:47:05,760 want to have a family and raise kids and things like that. 655 00:47:06,120 --> 00:47:08,720 You know, that that that was part of it. 656 00:47:09,880 --> 00:47:14,960 And then the other part of it. And I forget the exact sequence 657 00:47:14,960 --> 00:47:21,160 of all of this was on one of my trips to Rolls Royce, I was 658 00:47:21,400 --> 00:47:27,240 called in to see the chief engineer who said, oh, by the 659 00:47:27,240 --> 00:47:30,520 way, please let us know whenever you want to come back to this 660 00:47:30,520 --> 00:47:35,520 country and we'll sort it. So I had that standing offer 661 00:47:35,520 --> 00:47:41,320 from Rolls Royce that they would organise it for me or that I 662 00:47:41,320 --> 00:47:43,560 could basically choose where I came back to. 663 00:47:45,080 --> 00:47:51,280 So yeah, so I decided, yes, I wanted to, to, to come home and, 664 00:47:51,520 --> 00:47:57,920 and, and I chose Oxford rather than Cambridge as the place to 665 00:47:57,920 --> 00:48:03,000 come to partly. Well, you see in, in Cambridge 666 00:48:03,000 --> 00:48:05,640 I'd have probably been in engineering, not in maths. 667 00:48:06,680 --> 00:48:12,520 You know, Cambridge maths was always kind of anti numerical 668 00:48:12,520 --> 00:48:18,480 analysis. There was certainly a chunk of 669 00:48:18,480 --> 00:48:23,000 the faculty who felt that you know the computer is what you 670 00:48:23,000 --> 00:48:25,360 used if you weren't clever enough to do it properly. 671 00:48:26,680 --> 00:48:36,640 Old school. Old school, yes, whereas Oxford 672 00:48:36,880 --> 00:48:40,840 really had had embraced numerical methods and had very, 673 00:48:40,840 --> 00:48:45,400 very strong group here and in those days the numerical 674 00:48:45,400 --> 00:48:48,680 analysis group, although they were a group of mathematicians, 675 00:48:49,000 --> 00:48:51,080 they were in the computer science department. 676 00:48:51,400 --> 00:48:54,320 And so there were also people in computer science on the parallel 677 00:48:54,320 --> 00:48:56,720 computing site. And so that that was an 678 00:48:56,720 --> 00:49:00,160 attraction to me as well, you know, to have them as sort of 679 00:49:00,160 --> 00:49:03,760 neighbors. As it turned out that that 680 00:49:03,760 --> 00:49:08,320 didn't work out because soon after arriving in 92, it was 681 00:49:08,320 --> 00:49:13,400 thefirst.com boom and all, all of the parallel computing people 682 00:49:13,560 --> 00:49:19,280 left to set up companies. So anyway, but but that was part 683 00:49:19,280 --> 00:49:21,240 of the motivation of choosing Oxford. 684 00:49:21,240 --> 00:49:25,040 Ah, OK. So yeah. 685 00:49:25,040 --> 00:49:27,800 So, you know, Sir, Rolls Royce kind of organized it. 686 00:49:28,560 --> 00:49:33,520 I mean it, it had to be a properly advertised and competed 687 00:49:33,520 --> 00:49:36,040 for a position. So I was in this strange 688 00:49:36,040 --> 00:49:40,680 position of helping to write the job description, possibly even 689 00:49:40,680 --> 00:49:44,560 the advert for a position that I then applied for. 690 00:49:44,680 --> 00:49:50,040 You know, there are various parts of my career which I, I, I 691 00:49:50,040 --> 00:49:52,480 look back on. And now, now I realise just how 692 00:49:52,480 --> 00:49:56,520 peculiar they were at the time. That's right. 693 00:49:56,560 --> 00:50:01,360 So you technically came into the computer science department? 694 00:50:01,400 --> 00:50:03,080 Yeah. So in those days it was called 695 00:50:03,080 --> 00:50:05,440 the Computing Laboratory. Yeah. 696 00:50:05,840 --> 00:50:08,600 And that was a Keble Rd. always OK. 697 00:50:09,640 --> 00:50:13,120 And so that was like a Rolls Royce. 698 00:50:13,800 --> 00:50:18,760 So it was a Rolls Royce readership in CFD and then and 699 00:50:18,760 --> 00:50:21,760 then they funded me to set up a whole research group. 700 00:50:21,760 --> 00:50:25,560 So I, I have a lot of funding from them for a, you know, 701 00:50:25,560 --> 00:50:31,880 prolonged period. So really from 92 through to 702 00:50:32,040 --> 00:50:37,320 2008 is when I moved to maths, you know, so, so, so for those 703 00:50:37,320 --> 00:50:42,240 sort of 15 years, you know, I, I had a lot of Rolls Royce 704 00:50:42,240 --> 00:50:47,280 funding. So this is the bit that I was 705 00:50:47,600 --> 00:50:49,080 wondering about. So you came into computer 706 00:50:49,080 --> 00:50:57,000 science for CFD and is this the time then when I guess what most 707 00:50:57,000 --> 00:51:00,920 people recognize as the Hydra code is, is this where it begins 708 00:51:00,920 --> 00:51:03,440 or did it actually begin even when you were MIT? 709 00:51:03,440 --> 00:51:06,440 When, when would when would you say was the start of that? 710 00:51:07,440 --> 00:51:10,880 Journey. So I would say that Hydra proper 711 00:51:10,880 --> 00:51:18,240 started in about 96, but there was there were various bits of 712 00:51:18,240 --> 00:51:22,000 research that in the sense laid the foundations for Hydra. 713 00:51:22,840 --> 00:51:30,200 So while I was still at MIT, you know, so I was thinking about 714 00:51:30,400 --> 00:51:36,400 what to do after onsflow. And so my plan after onsflow was 715 00:51:36,400 --> 00:51:43,400 that we really wanted something that would be a design tool for 716 00:51:43,400 --> 00:51:48,560 doing complete engines, you know, so no no longer single 717 00:51:48,560 --> 00:51:50,640 stage, really looking multi stage. 718 00:51:53,400 --> 00:51:58,600 I was also interested in the idea that the unsteadiness could 719 00:51:58,600 --> 00:52:01,240 be done from a linear perturbation point of view 720 00:52:01,520 --> 00:52:03,640 rather than doing non linear unsteady. 721 00:52:04,760 --> 00:52:10,200 So so also linearize the unsteady equations, look at, you 722 00:52:10,200 --> 00:52:13,960 know, harmonics. And so doing this for both 723 00:52:13,960 --> 00:52:19,320 flutter and forced response. The key technical issue there is 724 00:52:19,320 --> 00:52:25,720 whether it was legitimate to do linearised harmonic analysis of 725 00:52:25,720 --> 00:52:31,600 shock capturing. And so I had a student who did a 726 00:52:31,600 --> 00:52:35,560 project at MIT to prove that, show that that was was a 727 00:52:35,560 --> 00:52:39,960 legitimate thing to do that as long as you so slightly smeared 728 00:52:39,960 --> 00:52:44,160 the shock over a few grid points that that the linearised 729 00:52:44,160 --> 00:52:49,800 analysis did do the right thing and did, yeah, you know, you 730 00:52:49,800 --> 00:52:54,920 could do it on that basis. So that was the precursor work 731 00:52:55,680 --> 00:53:02,600 at MIT. And then early on in Oxford, I 732 00:53:02,600 --> 00:53:06,080 don't think I'd started this coding before I moved. 733 00:53:07,840 --> 00:53:13,360 I did a code called Slick SLIQ. So steady, linear and quadratic. 734 00:53:13,880 --> 00:53:17,440 So the idea was you did you know the non linear steady state, you 735 00:53:17,440 --> 00:53:21,720 did a linear perturbation analysis for the unsteady 736 00:53:21,720 --> 00:53:26,320 effects and then the quadratic was to it's kind of a formal 737 00:53:26,320 --> 00:53:31,760 asymptotic expansion to get the mean flow changes due to the 738 00:53:31,760 --> 00:53:39,720 second order quadratic effects. So, yeah, so I developed slick 739 00:53:39,720 --> 00:53:45,920 early in, in in my Oxford days and and you know, I had a 740 00:53:45,920 --> 00:53:48,640 student who who, who worked on that with with me. 741 00:53:50,120 --> 00:53:56,960 So that was one piece of work. Another piece of work was back 742 00:53:56,960 --> 00:54:02,880 in the 9293 era, there was funding. 743 00:54:02,880 --> 00:54:06,680 I think this came from, you know, the UK Department of Trade 744 00:54:06,680 --> 00:54:10,600 and Industry DTI. In those days there was an 745 00:54:10,600 --> 00:54:14,560 initiative of setting up parallel application centres in 746 00:54:14,560 --> 00:54:20,400 various parts of the country. And so I, I had a colleague on 747 00:54:20,400 --> 00:54:24,040 the computer science side in Oxford, Bill McCall, who'd 748 00:54:24,040 --> 00:54:26,880 applied for that funding even before I arrived. 749 00:54:27,360 --> 00:54:34,960 And so had funding for an IBM machine, something called an SP2 750 00:54:36,720 --> 00:54:43,840 and and also there was funding there for, for research and also 751 00:54:43,840 --> 00:54:46,080 also match funding. I think so. 752 00:54:46,240 --> 00:54:51,160 So I had 5050 matching from Rolls Royce and so did the 753 00:54:51,160 --> 00:54:58,480 project there on developing. So in the sense of support layer 754 00:54:58,480 --> 00:55:01,160 for doing distributed memory parallel computing. 755 00:55:01,560 --> 00:55:04,760 So this was something called, you know, we called OO plus 756 00:55:04,960 --> 00:55:07,840 Oxford parallel library for unstructured solvers. 757 00:55:10,000 --> 00:55:15,800 Although although orally O plus doesn't doesn't sound right, but 758 00:55:15,800 --> 00:55:21,280 written down it looks good. So, yeah, so, so I think Bill, 759 00:55:21,760 --> 00:55:26,680 yeah, must have been in Oxford after I arrived from maybe two 760 00:55:26,680 --> 00:55:30,560 or three years before he left in that.com boom. 761 00:55:31,480 --> 00:55:34,600 He was one of the people I hoped to work, worked with a bit more 762 00:55:34,600 --> 00:55:39,520 on the parallel computing side. So, so we got a whole pile of 763 00:55:39,520 --> 00:55:44,880 funding half and Rolls Royce half from DTI to develop this 764 00:55:44,880 --> 00:55:53,320 parallel application framework. There was another code that at 765 00:55:53,320 --> 00:55:58,160 the postdoc Paul Crompton, who actually did most of the 766 00:55:58,160 --> 00:56:03,520 development of that software and wrote the CFD code as basically 767 00:56:03,520 --> 00:56:06,760 a test bed to check everything worked correctly. 768 00:56:07,920 --> 00:56:11,520 But that was never intended as ACFD code for Rolls Royce. 769 00:56:14,920 --> 00:56:17,920 So then that's what led into Hydra. 770 00:56:18,320 --> 00:56:23,200 So basically there were the the ideas of doing steady post 771 00:56:23,200 --> 00:56:28,120 linear perturbation tested out in slick and there was the 772 00:56:28,280 --> 00:56:32,720 parallel framework in O plus that we'd developed thoroughly 773 00:56:33,240 --> 00:56:38,280 tested out. And so Hydra then what was built 774 00:56:38,280 --> 00:56:42,720 on those foundations now. So Hydra it dropped the idea of 775 00:56:42,720 --> 00:56:47,400 doing the quadratic piece. So, so it's non linear steady, 776 00:56:47,400 --> 00:56:53,240 or at least in its original incarnation, non linear, steady 777 00:56:53,920 --> 00:57:01,680 linear perturbation for flutter forced response and then 778 00:57:01,760 --> 00:57:07,200 adjoints of all of those for for the design optimization, both 779 00:57:07,200 --> 00:57:10,480 steady and and the unsteady aspects, you know, so the 780 00:57:10,480 --> 00:57:15,480 adjoints was completely motivated by the work that 781 00:57:15,480 --> 00:57:19,520 Anthony Jamieson was doing, you know, on on, you know, the 782 00:57:19,520 --> 00:57:27,080 aircraft side, except that I chose a different technical 783 00:57:27,080 --> 00:57:32,320 approach in as much as Anthony always viewed the adjoint being 784 00:57:32,440 --> 00:57:36,920 developed at the PDE level, you know, formulating the adjoint 785 00:57:36,920 --> 00:57:41,200 PDE and then thinking about how to discretize it. 786 00:57:41,720 --> 00:57:45,400 Whereas I followed the the so-called discrete adjoint 787 00:57:45,400 --> 00:57:51,560 approach, where where you take the non linear discrete 788 00:57:51,560 --> 00:57:56,360 equations, you linearize those sort of element by element and 789 00:57:56,360 --> 00:58:00,280 then you take the transpose of the matrix to define the the 790 00:58:00,280 --> 00:58:02,880 discrete adjoint. You know and. 791 00:58:04,400 --> 00:58:06,840 Yes. How did you, just out of 792 00:58:06,840 --> 00:58:11,600 interest at this time, maybe also MIT, but Oxford, how active 793 00:58:11,600 --> 00:58:17,640 were you in, you know, the AI AA, the the sort of turbo 794 00:58:17,640 --> 00:58:23,920 machinery conferences were you, were you sort of always, were 795 00:58:23,920 --> 00:58:27,320 you at these events and saw these people or was this more of 796 00:58:27,440 --> 00:58:31,080 a like direct industrial engagement? 797 00:58:31,080 --> 00:58:33,720 I'm always interested, like now I go to the AI AA conference. 798 00:58:33,720 --> 00:58:36,120 I'm always, just always wondering what it was like, you 799 00:58:36,120 --> 00:58:40,800 know, before and was that still the main venue I guess to go. 800 00:58:41,640 --> 00:58:44,720 Yes, yes, that, that, that, that was the main venue. 801 00:58:44,920 --> 00:58:53,760 I, I sometimes went to the ASMEIGTI conference, but I went 802 00:58:53,760 --> 00:58:57,200 more, more to the AI AA conferences. 803 00:58:57,240 --> 00:59:03,240 You know, I think there was more discussion of CFD at at AI AA 804 00:59:03,240 --> 00:59:06,240 and in, in, in particular, you know, the AI double ACFD 805 00:59:06,240 --> 00:59:09,680 conference. I mean that that was really my 806 00:59:09,680 --> 00:59:13,160 home, you know, sort of during this period. 807 00:59:13,160 --> 00:59:19,200 I would say that, you know, the ASME was more on the application 808 00:59:19,200 --> 00:59:23,360 side. Yeah. 809 00:59:23,360 --> 00:59:26,800 So I, I, I, I can remember going to two or three of those, but it 810 00:59:26,800 --> 00:59:29,240 was the AI AA ones which really were. 811 00:59:29,240 --> 00:59:30,880 Where was was this Reno? No. 812 00:59:30,880 --> 00:59:36,080 Where was the CFD? So Reno was the January 813 00:59:36,080 --> 00:59:39,480 conference, the CFD conference, what was in the summer. 814 00:59:40,000 --> 00:59:47,600 So I I remember well, I remember one in Snowmass in, in Colorado. 815 00:59:48,000 --> 00:59:49,360 Yeah. Beautiful venue. 816 00:59:49,360 --> 00:59:50,240 Yeah. Good, good. 817 00:59:50,440 --> 00:59:54,640 Orienteering, right? Well, yeah, yeah, except it's 818 00:59:54,640 --> 00:59:57,280 high, high enough up that you really wouldn't be wanting to 819 00:59:57,280 --> 00:59:59,440 run at that altitude. No, Yeah, yeah. 820 00:59:59,440 --> 01:00:02,040 I mean, this is this is a ski resort that in the in the 821 01:00:02,040 --> 01:00:05,240 summer. You know, they, you know, are 822 01:00:05,480 --> 01:00:07,720 used for conferences because there aren't that many people 823 01:00:07,720 --> 01:00:13,360 who want to go hiking. But I remember Snowmass. 824 01:00:13,440 --> 01:00:20,640 I remember Hawaii. I remember there was 1 in LA. 825 01:00:21,520 --> 01:00:23,320 Yeah. I mean, it just hopped, hopped 826 01:00:23,320 --> 01:00:26,400 all over the place. Yeah. 827 01:00:26,400 --> 01:00:27,560 So. Yeah. 828 01:00:27,560 --> 01:00:33,040 So I was really mainly in those days going to engineering 829 01:00:33,040 --> 01:00:39,760 conferences, not so many, I guess some maths conferences, 830 01:00:39,760 --> 01:00:45,160 but, but in those days, yeah, most of my publishing was in in 831 01:00:45,160 --> 01:00:47,240 engineering journal still at that point. 832 01:00:47,240 --> 01:00:49,840 But did you struggle? I'm interested because I, 833 01:00:51,000 --> 01:00:53,440 because you have such a strong maths background, did you ever 834 01:00:53,440 --> 01:00:58,320 struggle getting accepted or being in that blur of what 835 01:00:58,320 --> 01:01:02,520 journal, what conference is the right place for maths? 836 01:01:02,520 --> 01:01:05,040 And then I always find this interesting that, you know, it's 837 01:01:05,040 --> 01:01:07,160 something too applied or too fundamental. 838 01:01:07,160 --> 01:01:09,480 Did you find a sweet spot or was there still a little bit of a 839 01:01:09,480 --> 01:01:14,520 straight frustration that your deep maths wasn't understood by 840 01:01:14,520 --> 01:01:16,200 everybody? Do do you know what I'm getting 841 01:01:16,200 --> 01:01:23,080 at or was it not an issue then? I mean, I, I think I was 842 01:01:23,080 --> 01:01:27,840 probably in those days still viewed more as an engineer than 843 01:01:27,880 --> 01:01:33,160 than the mathematics, you know. I mean, you know, if you have a 844 01:01:33,560 --> 01:01:37,680 PhD in in aeronautics from MIT, you know, you're, you're an 845 01:01:37,680 --> 01:01:38,360 engineer. Yeah. 846 01:01:39,800 --> 01:01:42,760 So, I mean, there was never a question of the engineering 847 01:01:42,760 --> 01:01:48,440 community not accepting me. You know, I think, you know, my, 848 01:01:48,440 --> 01:01:51,800 my evolution has, has been one of the of the mass community 849 01:01:51,800 --> 01:01:52,920 accepting me. OK, OK. 850 01:01:55,520 --> 01:01:57,640 Yeah. So, yeah. 851 01:01:57,640 --> 01:02:02,360 And that, that, that I guess, you know, happened more once 852 01:02:02,560 --> 01:02:05,120 once I moved over into the Maths Institute. 853 01:02:06,440 --> 01:02:13,360 But yes, I guess I felt, I mean, having made the, the move to the 854 01:02:13,360 --> 01:02:18,520 numerical analysis group in Oxford, then in a sense I felt I 855 01:02:18,520 --> 01:02:23,160 was back amongst mathematicians. But but I was still very much at 856 01:02:23,160 --> 01:02:26,200 the engineering end of of of the group. 857 01:02:26,600 --> 01:02:33,200 You know, I think I was probably 40 before I wrote my first paper 858 01:02:33,200 --> 01:02:39,160 that had a theorem and a proof in it, you know, So I mean that 859 01:02:39,200 --> 01:02:41,440 that's kind of needed to be a mathematician. 860 01:02:42,600 --> 01:02:45,040 So what was the link to the engineering department at that 861 01:02:45,040 --> 01:02:50,120 time like because there's was, was it the case that that was 862 01:02:50,120 --> 01:02:55,680 more experimental work and the sort of CFD was mainly computer 863 01:02:55,680 --> 01:02:58,480 science? Is, is that sort of, I always 864 01:02:58,480 --> 01:03:01,160 found that unique in Oxford that, you know, like I said, MIT 865 01:03:01,160 --> 01:03:05,280 engineering where like, yeah, I guess Oxford's a bit different. 866 01:03:05,960 --> 01:03:11,200 So in in Oxford at that time we had three UT CS. 867 01:03:11,400 --> 01:03:14,920 So the UT CS are the university technology centres that, that 868 01:03:14,920 --> 01:03:22,840 the Rolls Royce set up. And so I had mine in CFD, there 869 01:03:22,840 --> 01:03:30,160 was an experimental one in heat transfer in, in, in engineering 870 01:03:30,160 --> 01:03:33,040 science. And then there was one in 871 01:03:33,040 --> 01:03:40,560 materials which I didn't didn't have any interaction with except 872 01:03:40,840 --> 01:03:43,120 one of the profs. There used to be an orienteer 873 01:03:43,120 --> 01:03:44,840 back in my case. So there was. 874 01:03:46,200 --> 01:03:51,280 A UTC just for CFD. Yes, yes, yes. 875 01:03:51,280 --> 01:03:55,560 So this was my, my own little UTC. 876 01:03:55,560 --> 01:04:03,000 I was the only academic in it. So and, and possibly within the 877 01:04:03,000 --> 01:04:07,040 Rolls Royce family, that's slightly unusual to have a UTC 878 01:04:07,040 --> 01:04:12,720 that only has one academic. Generally they're they're bigger 879 01:04:12,720 --> 01:04:15,920 than that. But it was a funding mechanism, 880 01:04:15,920 --> 01:04:17,680 I guess, and you were doing the work. 881 01:04:18,600 --> 01:04:22,840 Yes, the funding mechanism it it, it involved me in all the 882 01:04:22,840 --> 01:04:26,040 UTC directors meetings. I mean, you were part of the 883 01:04:26,040 --> 01:04:31,080 family. And I mean that that's kind of 884 01:04:31,080 --> 01:04:35,640 important in the sense that Rolls Royce really knew how to 885 01:04:35,640 --> 01:04:43,680 work well with academics, that you were part of the family in 886 01:04:43,680 --> 01:04:46,840 the sense that you knew all the problems as well as, you know, 887 01:04:46,840 --> 01:04:49,440 the, the achievements of, of, of Rolls Royce. 888 01:04:49,440 --> 01:04:54,040 You know, they, they, they, you know, they didn't hide anything 889 01:04:54,040 --> 01:04:57,280 from you so that you could think about what you might potentially 890 01:04:57,280 --> 01:05:01,360 do to, to, to help them address some of their challenges and 891 01:05:01,360 --> 01:05:06,560 things. So, you know, it really was a 892 01:05:06,560 --> 01:05:13,040 very good collaborative experience when, when you're 893 01:05:13,040 --> 01:05:16,320 working with industry, you know, it's important that both sides 894 01:05:16,320 --> 01:05:20,200 realise that what the other wants out of the relationship 895 01:05:20,200 --> 01:05:23,440 is, is different, you know, and, and, and so you're always 896 01:05:23,440 --> 01:05:26,760 looking for this sort of win win arrangement, you know, the, you 897 01:05:26,760 --> 01:05:29,720 know, so they understood that, you know, for us, it was 898 01:05:29,720 --> 01:05:33,680 important for the students to publish papers, to write 899 01:05:33,760 --> 01:05:37,080 dissertations, you know, that there would be times when they 900 01:05:37,080 --> 01:05:39,880 would be utterly focused on writing their dissertation and 901 01:05:40,160 --> 01:05:44,800 not doing any more research, you know, but equally, I, I 902 01:05:44,800 --> 01:05:48,280 understood what Rolls Royce needed out of the relationship, 903 01:05:49,200 --> 01:05:53,120 which was primarily software, but occasionally if, if there 904 01:05:53,120 --> 01:05:56,840 was a particular engineering thing to be investigated, you 905 01:05:56,840 --> 01:06:01,840 know, we may occasionally do some, no special calculations 906 01:06:01,840 --> 01:06:05,520 just for them, as it were, rather than as, as part of the 907 01:06:05,520 --> 01:06:07,680 research. Yeah, I always found that 908 01:06:07,680 --> 01:06:13,280 interesting that some it takes a special company to understand 909 01:06:13,280 --> 01:06:17,000 the value of academic engagement and have the patience and the 910 01:06:17,000 --> 01:06:20,760 long term vision that it's not just cheap labour. 911 01:06:21,440 --> 01:06:26,040 Yes, you know that requires you. You have to be very careful to 912 01:06:26,040 --> 01:06:27,880 make sure it's never just cheap labour. 913 01:06:30,920 --> 01:06:34,520 I think Rolls Royce was possibly slightly disappointed that they 914 01:06:34,520 --> 01:06:38,360 never ended up being able to hire any of my students. 915 01:06:38,360 --> 01:06:41,680 You know, that's, that's, that's the other thing that, you know, 916 01:06:41,920 --> 01:06:47,240 Rolls Royce would ideally like from a UTC is as a source of, 917 01:06:47,440 --> 01:06:51,480 you know, people to be employed, you know, and that that never 918 01:06:51,480 --> 01:06:56,440 quite happened. But, yeah. 919 01:06:56,640 --> 01:06:59,280 But certainly they, they, they, they got their money's worth in 920 01:06:59,280 --> 01:07:04,560 terms of CFD codes. So, yeah, and I guess like in 921 01:07:04,560 --> 01:07:08,640 today's world probably, you know, you'd create a start up or 922 01:07:08,640 --> 01:07:12,800 something, write a code, you know, I guess at that time there 923 01:07:12,800 --> 01:07:17,520 was a more traditional link to the to the company, right? 924 01:07:17,520 --> 01:07:18,640 You know, in terms of. Yeah. 925 01:07:18,840 --> 01:07:22,600 I mean in, in, in this area, it would be tough to do a start up. 926 01:07:22,640 --> 01:07:27,520 I mean, I guess 11 Alonso's done, done the startup. 927 01:07:27,520 --> 01:07:30,880 I haven't talked to him recently as to how, how that's going. 928 01:07:32,240 --> 01:07:38,240 I mean it, yeah. At, at, at one point in parallel 929 01:07:38,240 --> 01:07:43,120 computing, I, you know, once I moved out of CFD into into 930 01:07:43,120 --> 01:07:47,560 mathematical finance, I, I tried setting up a spin off. 931 01:07:48,760 --> 01:07:52,280 No, it's just more hard work. Yeah, I'm, I'm, I'm, I'm 932 01:07:52,320 --> 01:07:56,520 fundamentally an academic, not, not a start up guy. 933 01:07:57,440 --> 01:08:01,960 The reason I say it is because, you know, speaking now to 934 01:08:01,960 --> 01:08:08,400 yourself or, or to Anthony, and I guess that like now CFD has 935 01:08:08,400 --> 01:08:13,160 become very dominated by these huge multibillion dollar 936 01:08:13,560 --> 01:08:18,720 commercial companies. I guess in the 80s, that was 937 01:08:18,920 --> 01:08:23,200 before the time, wasn't it? It was before the, the Fluence 938 01:08:23,200 --> 01:08:27,279 and, and, and the, the open phones and it was still, you 939 01:08:27,279 --> 01:08:31,600 wrote your own code. I assume that's old. 940 01:08:32,359 --> 01:08:35,040 Was that one of the reasons for the hydrogen development that 941 01:08:35,040 --> 01:08:37,479 they couldn't just buy something off the shelf that wasn't a 942 01:08:37,479 --> 01:08:40,120 company that could just sell them a capability they felt they 943 01:08:40,120 --> 01:08:41,160 needed? Yes. 944 01:08:41,160 --> 01:08:44,520 And even today I don't think there's a company that could 945 01:08:44,520 --> 01:08:48,000 sell them what they need because the turbo machinery requirements 946 01:08:48,479 --> 01:08:53,520 are really very specific. So I mean I've I've not kept up 947 01:08:53,520 --> 01:08:56,840 with the discipline. So I don't know what ANSYS 948 01:08:56,840 --> 01:09:03,479 Fluent has as a capability these days, but you know, the ability, 949 01:09:03,479 --> 01:09:09,560 for example, to have flutter calculations being performed on 950 01:09:09,560 --> 01:09:13,560 a single blade passage with an inter blade phase angle between 951 01:09:13,560 --> 01:09:16,359 the passages. You know, that's such a unique 952 01:09:16,560 --> 01:09:21,319 requirement of turbo machinery that I'm not sure the answers 953 01:09:21,319 --> 01:09:24,240 views the market as being big enough to develop that 954 01:09:24,240 --> 01:09:26,800 capability. And then, and then you get into 955 01:09:26,800 --> 01:09:30,680 things like real gas effects, you know, so, so, so Hydra 956 01:09:31,040 --> 01:09:34,040 doesn't assume a fixed gamma. You know, it, it, it has a 957 01:09:34,040 --> 01:09:38,319 general, you know, energy temperature relationship in 958 01:09:38,319 --> 01:09:43,000 there. So again, I mean, I guess things 959 01:09:43,000 --> 01:09:46,960 like that answers could add in, but there will be add insurance 960 01:09:46,960 --> 01:09:50,399 for particular customers and, and they would charge 961 01:09:50,399 --> 01:09:54,720 accordingly. I mean, one of the reasons I got 962 01:09:54,720 --> 01:10:01,480 out of the CFD business is it's not clear to me long term, you 963 01:10:01,480 --> 01:10:06,480 know, what Rolls Royce will do for their next so CFD code. 964 01:10:06,920 --> 01:10:12,040 So I mean, Hydra's just celebrated its 25th anniversary 965 01:10:12,040 --> 01:10:17,600 at, at, at Rolls Royce. I would say it'll remain the 966 01:10:17,600 --> 01:10:21,920 corporate, you know, primary code for at least another 10 967 01:10:21,920 --> 01:10:25,360 years because it takes a long time to, to introduce a brand 968 01:10:25,360 --> 01:10:29,160 new code. And you know, they, you know, 969 01:10:29,160 --> 01:10:36,920 they're doing work with Spencer Sherwin on his Nectar code with 970 01:10:36,920 --> 01:10:41,400 the thought that they will use that in, in applications where 971 01:10:41,400 --> 01:10:47,600 they want to really do DNS or or at least high resolution LES, 972 01:10:48,120 --> 01:10:53,520 but, but that won't be part of their sort of standard, you 973 01:10:53,520 --> 01:11:00,600 know, design process. So I think it's, it's harder to 974 01:11:00,600 --> 01:11:06,040 see now. Yeah, companies funding a brand 975 01:11:06,040 --> 01:11:10,240 new code development. That's kind of what I was 976 01:11:10,280 --> 01:11:15,480 hinting at, that it does seem as if that that generation in the 977 01:11:15,480 --> 01:11:21,760 80s and the 90s throughout aerospace was a time of real 978 01:11:21,760 --> 01:11:23,840 innovation and development. It was. 979 01:11:23,840 --> 01:11:27,560 Themselves. Yeah, I mean when, when, when I, 980 01:11:27,880 --> 01:11:32,280 you know, joined Rolls Royce's first CFD group in 1980. 981 01:11:32,680 --> 01:11:36,520 You know, most of the design methods were one-dimensional 982 01:11:36,520 --> 01:11:42,720 design methods, you know, with sort of mean, mean line, sort 983 01:11:42,720 --> 01:11:46,560 of, you know, I mean really mathematical models rather than 984 01:11:46,560 --> 01:11:51,760 numerics. So, yeah, I mean it, it, it's 985 01:11:51,760 --> 01:11:57,240 been wonderful just being part of that whole process of, you 986 01:11:57,240 --> 01:12:01,880 know, you know, the development of computational engineering 987 01:12:01,880 --> 01:12:07,680 and, and seeing it go all the way from 1D2D3D, you know, in 988 01:12:07,680 --> 01:12:12,040 viscid to viscous steady to unsteady and then the adjoints 989 01:12:12,040 --> 01:12:16,160 and everything. You know, it is interesting. 990 01:12:16,160 --> 01:12:21,120 So looking back over your lifetime and looking at the 991 01:12:21,120 --> 01:12:24,360 progress in computational engineering and then of course 992 01:12:24,360 --> 01:12:27,040 the progress in parallel computing and the high 993 01:12:27,040 --> 01:12:30,080 performance computing, I mean. Both of them, yeah. 994 01:12:30,360 --> 01:12:33,040 It's really stunning looking back over the years. 995 01:12:34,000 --> 01:12:38,720 Yeah, that's what I was wondering about is so you, so 996 01:12:38,720 --> 01:12:40,920 you came into Oxford doing the computer science. 997 01:12:42,280 --> 01:12:45,600 I was just looking at the and I guess this is where we have a 998 01:12:45,600 --> 01:12:48,600 slight shared connection into some of the people. 999 01:12:51,240 --> 01:12:54,720 How did and you mentioned also about the government setting up 1000 01:12:54,720 --> 01:12:56,440 some, you know, parallel computing. 1001 01:12:57,400 --> 01:13:02,920 I guess how did that evolve the HPC angle, Oxford, maybe some of 1002 01:13:02,920 --> 01:13:06,080 the E research ideas you'll move into mass. 1003 01:13:06,080 --> 01:13:09,360 When did how did that time? And I guess that probably aligns 1004 01:13:09,360 --> 01:13:14,280 to also when you had a declining interest in CFD and more into 1005 01:13:14,280 --> 01:13:17,120 the computing and other areas. Is that fair to say? 1006 01:13:17,600 --> 01:13:20,280 Were they aligned a little bit or were they sort of separate? 1007 01:13:21,000 --> 01:13:25,160 OK, so yeah, yeah, let's let let's. 1008 01:13:25,480 --> 01:13:26,800 Try to. I've asked you like 3 questions 1009 01:13:26,800 --> 01:13:29,320 in one. Yeah, yeah, yeah. 1010 01:13:29,360 --> 01:13:32,280 So get get the timeline straight in my mind. 1011 01:13:32,880 --> 01:13:36,240 OK. So the main Hydra research 1012 01:13:36,240 --> 01:13:47,440 period was sort of 96 to maybe 2004, 2006, something like that. 1013 01:13:48,840 --> 01:13:55,920 And the Hydra code was built on top of this O plus parallel 1014 01:13:55,920 --> 01:13:59,360 layer that we had developed earlier as part of the DTI 1015 01:13:59,360 --> 01:14:05,480 funded activity. And so incidentally, both O plus 1016 01:14:05,480 --> 01:14:09,480 and Hydra, you're the IPR belongs to Rolls Royce. 1017 01:14:12,480 --> 01:14:20,960 So O plus was very much developed in the time of single 1018 01:14:20,960 --> 01:14:27,280 core CPUs, risk risk based CPUs, you know, so, so there, you 1019 01:14:27,280 --> 01:14:30,160 know, there was good performance there, but single core. 1020 01:14:30,840 --> 01:14:35,000 And then what was happening over time was that CPUs were, were 1021 01:14:35,000 --> 01:14:41,920 going multicore and, and then we had GPUs that came along. 1022 01:14:42,840 --> 01:14:47,960 So, so again, I've, I've always been interested in the latest 1023 01:14:47,960 --> 01:14:53,920 computing technology. So I think it must have been 1024 01:14:56,320 --> 01:15:06,160 sort of late 2006, early 2007. An ex colleague, Mike Rodger, I 1025 01:15:06,160 --> 01:15:08,000 don't know if the name means anything to you. 1026 01:15:08,720 --> 01:15:16,120 He, he set up a company spin off from Warwick actually, which 1027 01:15:16,200 --> 01:15:22,040 sold parallel, you know, systems, you know, parallel 1028 01:15:22,040 --> 01:15:26,280 clusters. And he was the one who said to 1029 01:15:26,280 --> 01:15:32,720 me, there's this new thing called the GPU and this, this, 1030 01:15:32,720 --> 01:15:36,520 this new language CUDA. Actually, I'm not even sure if 1031 01:15:36,520 --> 01:15:38,840 he told me about CUDA. He may just have told me, you 1032 01:15:38,840 --> 01:15:41,000 know, there's this new hardware called GPUs. 1033 01:15:41,200 --> 01:15:43,720 It's really impressive. You know, you ought to have a 1034 01:15:43,720 --> 01:15:51,640 look at it. And at about the same time there 1035 01:15:51,640 --> 01:15:57,440 was a company based in Bristol called Clearspeed, which had 1036 01:15:57,440 --> 01:16:01,300 people in it who I think had the background coming from the in 1037 01:16:01,300 --> 01:16:07,720 MOS transputer days, you know, so, so I first tried the the 1038 01:16:07,720 --> 01:16:14,400 Clearspeed card and I was impressed by what it was capable 1039 01:16:14,400 --> 01:16:17,800 of, but I wasn't blown away by it. 1040 01:16:20,240 --> 01:16:27,320 And then actually I had the visiting student, a Chinese 1041 01:16:27,320 --> 01:16:31,760 student, I forget how, how he came to me, but he, he did the 1042 01:16:31,760 --> 01:16:35,040 clear speed work. And then I said, well, now, now 1043 01:16:35,040 --> 01:16:40,120 let's try this GPU. And he came back to me with it 1044 01:16:40,120 --> 01:16:45,240 within like a couple of weeks with the code performing just 1045 01:16:45,240 --> 01:16:48,120 incredibly fast. And I was saying, OK, you've 1046 01:16:48,120 --> 01:16:55,840 clearly messed up the timing. And so it took took me a week to 1047 01:16:55,840 --> 01:16:59,320 convince myself, no, I mean it really was performing that that 1048 01:16:59,320 --> 01:17:03,760 well. So, so this was right at the 1049 01:17:03,760 --> 01:17:08,840 beginning of CUDA. So I think we, we started with 1050 01:17:09,360 --> 01:17:12,440 the naughty .9 beta release or something like that. 1051 01:17:14,080 --> 01:17:18,120 And, and so I was just stunned at, you know, the power of, of 1052 01:17:18,120 --> 01:17:20,680 the GPU. Now this was when I was already 1053 01:17:20,680 --> 01:17:23,840 starting to pivot into mathematical finance. 1054 01:17:23,840 --> 01:17:27,320 So I was actually looking at it for doing Monte Carlo 1055 01:17:28,360 --> 01:17:34,240 simulations in, in finance, which is why NVIDIA then got 1056 01:17:34,240 --> 01:17:37,800 interested in me because they saw that as a potential market. 1057 01:17:40,320 --> 01:17:46,800 So, yeah, so, so I was pivoting to finance, but I was also 1058 01:17:46,800 --> 01:17:49,160 interested in the power of the GPUs. 1059 01:17:49,520 --> 01:17:53,680 So after two or three years of that, I'm trying to think 1060 01:17:53,680 --> 01:18:00,200 exactly on the timing. Again, I was interested in doing 1061 01:18:00,200 --> 01:18:07,240 an upgrade to O Plus which could then incorporate both GPUs and 1062 01:18:07,240 --> 01:18:12,240 multi core CPUs. But by the way, just before you 1063 01:18:12,240 --> 01:18:15,360 go on, did so did you if it was 2000 and six 2007, did you 1064 01:18:15,360 --> 01:18:19,200 interact with Ian Buck at all then on the cooter stuff or was 1065 01:18:19,200 --> 01:18:23,440 it on people more in the UK who? You no, no. 1066 01:18:23,440 --> 01:18:30,200 So Massimiliano I, I interacted with Ian. 1067 01:18:30,360 --> 01:18:37,200 Ian, I probably met in in the same way that I met Jensen at, 1068 01:18:37,320 --> 01:18:42,920 you know, drinks things at GTC. So, so I was, I think the number 1069 01:18:42,920 --> 01:18:45,120 2 CUDA fellow. I think. 1070 01:18:45,120 --> 01:18:48,920 I think the first CUDA fellow was somebody in India. 1071 01:18:49,760 --> 01:18:52,600 And then I was #2 like a month later. 1072 01:18:54,080 --> 01:18:59,480 So yeah, yeah, I mean, yeah. So I was going along to the GTC, 1073 01:18:59,600 --> 01:19:02,600 you know. I have to say to the, I guess to 1074 01:19:02,600 --> 01:19:06,360 the people is that a bit of like self-proclaimed interest just 1075 01:19:06,360 --> 01:19:09,640 because now the team I'm in has those people in it. 1076 01:19:09,640 --> 01:19:12,280 So I'm always intrigued when they're like, oh, you, you know 1077 01:19:12,280 --> 01:19:14,280 that person and this person knows this person. 1078 01:19:14,320 --> 01:19:17,480 So it's kind of interesting, especially because you were, I 1079 01:19:17,480 --> 01:19:19,480 mean, now everybody knows NVIDIA and GP us. 1080 01:19:19,560 --> 01:19:24,000 But 2006 seven was really early on, wasn't it? 1081 01:19:24,680 --> 01:19:27,200 Yeah, yeah. And, and in fact, some of the 1082 01:19:27,200 --> 01:19:32,760 people from Clear Speed moved to NVIDIA in part in influenced by 1083 01:19:32,760 --> 01:19:36,840 my feedback to them. Yeah. 1084 01:19:36,840 --> 01:19:41,200 They then saw the writing on the wall and moved, moved ship 1085 01:19:41,200 --> 01:19:46,440 accordingly. So, but yes, I don't think I, 1086 01:19:46,880 --> 01:19:49,440 yeah, I don't remember talking to Ian Buck at all. 1087 01:19:49,920 --> 01:19:55,320 Massimiliano was was was one of my early contacts in in in the 1088 01:19:55,320 --> 01:20:00,400 Bay Area. Yeah. 1089 01:20:01,080 --> 01:20:06,240 So, yeah, yes, I sort of came back then to do this upgrade 1090 01:20:06,440 --> 01:20:11,680 from O plus to to OP 2. So that, that was with Gehan, 1091 01:20:12,040 --> 01:20:16,040 you know, mutilation and Ishwan regularly that you, you, you 1092 01:20:16,440 --> 01:20:21,720 both of them. So, and so that, that had 1093 01:20:21,720 --> 01:20:27,320 funding from both Rolls Royce and, you know, the UK EPSURC, 1094 01:20:28,560 --> 01:20:36,880 you know, government funding. So that has in a sense, you 1095 01:20:36,880 --> 01:20:41,120 know, protected Hydra's future by, by up upgrading the, the, 1096 01:20:41,320 --> 01:20:45,240 the underlying computing hardness for, for, you know, 1097 01:20:45,240 --> 01:20:51,560 modern systems. But but I haven't really done 1098 01:20:51,920 --> 01:20:56,320 CFD as such since about 2/2/2000. 1099 01:20:56,760 --> 01:20:58,640 Seven, I think was the last time. 1100 01:20:59,200 --> 01:21:02,640 So what was, yeah, maybe that's the elephant in the room. 1101 01:21:04,120 --> 01:21:07,600 Why what, why was the, I mean, you sort of hinted towards it 1102 01:21:07,920 --> 01:21:13,600 earlier, but what was the reason to maybe move away from CFD and 1103 01:21:13,600 --> 01:21:16,240 move into the, you know, the financial side? 1104 01:21:16,440 --> 01:21:20,240 Was that the maths angle moving to the maths department sort of 1105 01:21:20,240 --> 01:21:23,320 just wanting new challenges? The computing side What? 1106 01:21:23,320 --> 01:21:26,200 What was the ingredients to that? 1107 01:21:34,200 --> 01:21:37,440 Yeah, I'm just, I'm just thinking what, what, what what 1108 01:21:37,440 --> 01:21:40,400 to say. Yeah, let's, let's let's be be 1109 01:21:40,400 --> 01:21:44,000 open about it because I have been been open about this, you 1110 01:21:44,000 --> 01:21:51,800 know, in, in various settings. I got, you know, more than burnt 1111 01:21:51,800 --> 01:21:56,960 out developing the hydrocode. You know that there were lots of 1112 01:21:56,960 --> 01:22:00,400 aspects of the hydrocode that worked very well, but they had 1113 01:22:00,520 --> 01:22:05,160 major problems early on with numerical stability which got me 1114 01:22:05,400 --> 01:22:11,320 horrendously stressed out to to the point of serious 1115 01:22:12,280 --> 01:22:16,360 consequences. So, you know, for about 6 months 1116 01:22:16,360 --> 01:22:18,560 it took to, to recover from all of that. 1117 01:22:20,600 --> 01:22:27,520 And I decided at that point that it's a got to a situation where 1118 01:22:28,040 --> 01:22:31,640 in a sense, too much perspiration and too little 1119 01:22:31,640 --> 01:22:35,160 inspiration. It, it, it's not being fun, you 1120 01:22:35,160 --> 01:22:42,560 know, managing a large software project is, is tiring, you know, 1121 01:22:42,600 --> 01:22:48,080 and, and, you know, I'd, yeah, I'd, I'd had enough. 1122 01:22:48,680 --> 01:22:55,720 So, you know, I sort of finished up the things that needed to be 1123 01:22:55,720 --> 01:22:59,800 finished up, transferred things to Rolls Royce and they've, 1124 01:23:00,040 --> 01:23:02,920 they've continued, you know, the development subsequently. 1125 01:23:05,000 --> 01:23:10,440 So I decided I had to had to do something fresh different. 1126 01:23:10,920 --> 01:23:15,920 I had to get out of big codes. Yeah. 1127 01:23:16,760 --> 01:23:25,560 So in, in engineering style, I, I thought, OK, where's the money 1128 01:23:25,560 --> 01:23:28,400 in terms of where, where's the research funding? 1129 01:23:28,840 --> 01:23:32,880 What do, what do people care about health and wealth? 1130 01:23:33,360 --> 01:23:38,880 So, so computational biology and computational finance, those, 1131 01:23:39,120 --> 01:23:42,760 those were the two areas I contemplated moving into. 1132 01:23:43,280 --> 01:23:48,840 And computational biology, I, I know no biology. 1133 01:23:48,840 --> 01:23:51,840 So I'd have been starting from Ground Zero on that. 1134 01:23:52,240 --> 01:23:57,200 And I also wasn't convinced there was the research funding 1135 01:23:57,200 --> 01:24:00,960 in in that area, although I was proved wrong on that, you know, 1136 01:24:01,040 --> 01:24:03,080 you know, there is actually plenty of funding there. 1137 01:24:04,240 --> 01:24:08,000 Computational finance, there was a very good mathematical finance 1138 01:24:08,000 --> 01:24:12,080 group in the maths department but they didn't have a lot of 1139 01:24:12,080 --> 01:24:15,960 numerical expertise. So, so for me that that was a 1140 01:24:15,960 --> 01:24:19,800 natural fit. And So what I did was I actually 1141 01:24:19,800 --> 01:24:25,360 transferred from computer science into maths to join the 1142 01:24:25,360 --> 01:24:30,000 mathematical finance group. And then a couple of years later 1143 01:24:30,160 --> 01:24:33,240 the rest of the numerical analysis group kind of moved, 1144 01:24:33,400 --> 01:24:37,520 moved across behind me into maths as well. 1145 01:24:40,480 --> 01:24:45,000 And so now, now I'm, I'm head of the, the numerical analysis 1146 01:24:45,000 --> 01:24:48,600 group now, you know, so, so in a sense in, in, in the last few 1147 01:24:48,600 --> 01:24:52,560 years I've sort of transferred back from the finance group into 1148 01:24:52,560 --> 01:24:56,920 the numerical analysis group. As which ultimately sort of came 1149 01:24:56,920 --> 01:25:01,800 with you in the fullness of time from the computer science into 1150 01:25:01,800 --> 01:25:02,960 the into the. Maths, yeah. 1151 01:25:02,960 --> 01:25:06,480 When when when maths got its new building, there was the 1152 01:25:06,480 --> 01:25:09,840 opportunity for the numerical analysis group to move and it 1153 01:25:09,840 --> 01:25:12,160 was a take it to leave it kind of opportunity. 1154 01:25:12,160 --> 01:25:14,600 And you know, they, they, they, they moved. 1155 01:25:14,600 --> 01:25:17,240 And for people listening, it's a lovely building. 1156 01:25:17,560 --> 01:25:19,280 It's a very nice. It's a very nice. 1157 01:25:20,760 --> 01:25:25,080 Building also, you know, when I joined computer science in 92, 1158 01:25:25,320 --> 01:25:29,360 the numerical analysis group was half of the whole department, 1159 01:25:29,760 --> 01:25:33,200 you know, Wow yeah. So what happened over time was 1160 01:25:33,200 --> 01:25:36,280 the computer science side grew and the numerical analysis side 1161 01:25:36,280 --> 01:25:39,360 didn't. And so by by the time the 1162 01:25:39,360 --> 01:25:48,000 numerical analysis group moved over in 2910, yeah, they were 1163 01:25:48,000 --> 01:25:52,320 basically pushed out effectively, you know, you know, 1164 01:25:52,400 --> 01:25:56,200 you know, that it didn't make sense for them being in computer 1165 01:25:56,200 --> 01:25:59,560 science any longer. I mean, historically they were 1166 01:25:59,560 --> 01:26:03,360 there because that's where the computers were, you know, and, 1167 01:26:03,680 --> 01:26:07,160 and that, you know, that is a history that has happened 1168 01:26:07,160 --> 01:26:09,000 elsewhere. I mean, you know, numerical 1169 01:26:09,000 --> 01:26:12,200 analysis and Stanford for a long time was based in computer 1170 01:26:12,200 --> 01:26:14,400 science. And I think, again, 1171 01:26:14,400 --> 01:26:17,240 historically, it's because that's where the computers were. 1172 01:26:19,040 --> 01:26:20,920 I do always find this interesting and that's why I 1173 01:26:20,920 --> 01:26:24,480 just wanted to, you know, get the I get the quick history of 1174 01:26:25,320 --> 01:26:28,160 like places like the E Research Centre and others because in 1175 01:26:28,160 --> 01:26:32,720 some ways I understand the logic that where does some of these 1176 01:26:32,720 --> 01:26:34,840 places fit? You know, there's numerical 1177 01:26:34,840 --> 01:26:37,360 analysis, there's computers, there's high performance 1178 01:26:37,360 --> 01:26:39,600 computing, there's an engineering application, there's 1179 01:26:39,640 --> 01:26:46,600 a pure application was the was was that sort of E research 1180 01:26:46,600 --> 01:26:49,000 centre and I believe there were others around the country was 1181 01:26:49,000 --> 01:26:55,040 the initiative to try and bring them together to in a more 1182 01:26:55,040 --> 01:26:57,880 collaborative way Was that was that sort of the initiative? 1183 01:26:58,080 --> 01:27:02,560 I know that they've subsequently largely folded into other 1184 01:27:02,560 --> 01:27:05,400 departments now, but. Yes, I mean, it it there was a 1185 01:27:05,400 --> 01:27:13,880 massive funding initiative to fund this E science and yeah, 1186 01:27:13,880 --> 01:27:18,560 so, so Tony, hey was was the person in charge of that. 1187 01:27:21,440 --> 01:27:28,040 And I'm not sure what exactly the intent was. 1188 01:27:28,120 --> 01:27:33,360 How, how explicitly they wanted it to be an interdisciplinary 1189 01:27:33,360 --> 01:27:34,120 effort. Maybe. 1190 01:27:34,280 --> 01:27:40,440 Maybe they did so in Oxford. Yeah. 1191 01:27:41,280 --> 01:27:44,040 When, when, when they round up the usual suspects for high 1192 01:27:44,040 --> 01:27:46,720 performance computing. I was, I was one of the usual 1193 01:27:46,720 --> 01:27:49,320 suspects. And so I was one of the four 1194 01:27:49,320 --> 01:27:56,560 people who put in the Oxford bid, you know, to get OERC 1195 01:27:56,560 --> 01:28:04,080 initially, I mean, at that point and Trefethen was Tony's deputy 1196 01:28:04,080 --> 01:28:08,360 and then she later, you know, joined Oxford and and and became 1197 01:28:08,360 --> 01:28:15,360 head of OARC. So I think certainly computer 1198 01:28:15,360 --> 01:28:20,920 science in those days in Oxford was very theoretical, especially 1199 01:28:20,920 --> 01:28:24,640 after the panel computing people like Bill McColl had left 1200 01:28:24,640 --> 01:28:31,840 duringthe.com era to some extent in Oxford there were tensions 1201 01:28:31,840 --> 01:28:36,880 between engineering and computer science as to where some of the 1202 01:28:36,880 --> 01:28:41,120 more applied computer science activities should go, should 1203 01:28:41,120 --> 01:28:46,320 should go. And then OERC was, was just, 1204 01:28:46,400 --> 01:28:50,440 yeah, another, another location to have such things, you know, 1205 01:28:50,440 --> 01:28:54,560 so, so eventually it made sense for OOERC to be merged into 1206 01:28:54,560 --> 01:28:59,640 engineering. I think at the time computer 1207 01:28:59,640 --> 01:29:03,240 science expressed a strong view that they they did not want to 1208 01:29:03,240 --> 01:29:06,480 be the destination. So. 1209 01:29:06,760 --> 01:29:11,200 Anyway, OK, but maybe on to then more the maths side. 1210 01:29:11,240 --> 01:29:18,120 I mean, I'm, I'm intrigued because maybe to explain at more 1211 01:29:18,120 --> 01:29:22,560 of a higher level, if you can, what what are the similarities 1212 01:29:22,560 --> 01:29:26,280 between some of the mathematical finances and maybe CFD? 1213 01:29:26,280 --> 01:29:29,760 What are the, and obviously you're known for which, you 1214 01:29:30,640 --> 01:29:33,320 know, I only understand the basic basics of it, but the sort 1215 01:29:33,320 --> 01:29:36,120 of multi level Monte Carlo, which I as soon as I saw the 1216 01:29:36,120 --> 01:29:37,560 description, always like a multi grid. 1217 01:29:37,800 --> 01:29:40,400 It always makes me interested that there's some of these, you 1218 01:29:40,400 --> 01:29:44,000 see it with AI today that some of these CFD solutions to 1219 01:29:44,000 --> 01:29:46,440 problems are now being applied to new areas. 1220 01:29:46,680 --> 01:29:48,960 So what? Yeah, what are the sort of high 1221 01:29:48,960 --> 01:29:52,760 level things that makes the link between CFD and maths, I guess. 1222 01:29:53,480 --> 01:30:00,280 So in in mathematical finance, there's basically two kinds of 1223 01:30:00,280 --> 01:30:03,560 methodology. You can approach it from a PDE 1224 01:30:03,560 --> 01:30:08,360 point of view, where in in one sense you you have APDE that 1225 01:30:08,360 --> 01:30:12,720 describes the evolution of the probability density function 1226 01:30:13,080 --> 01:30:16,800 for, you know, a stock having a certain value at the time in the 1227 01:30:16,800 --> 01:30:19,920 future. And there's a corresponding sort 1228 01:30:19,920 --> 01:30:25,800 of adjoint of that to give the value of of various financial 1229 01:30:25,800 --> 01:30:29,840 options or there's the Monte Carlo approach where you 1230 01:30:29,840 --> 01:30:34,600 simulate lots of these different possible future trajectories of 1231 01:30:34,640 --> 01:30:39,680 of the stock and then and then say, OK, given that family of 1232 01:30:39,680 --> 01:30:43,200 solutions, what's the average pay off of your financial 1233 01:30:43,200 --> 01:30:48,240 option? So when I initially moved into 1234 01:30:48,240 --> 01:30:52,800 finance, I was focused on the PDE site because it's basically 1235 01:30:52,800 --> 01:30:56,320 convection diffusion Pdes. And I thought, hey, you know, 1236 01:30:58,120 --> 01:31:01,240 this is, this is a no brainer. I can just bring all the CFD 1237 01:31:01,240 --> 01:31:05,040 techniques over and, and, and do things here. 1238 01:31:07,080 --> 01:31:12,200 What I quickly found was other people have beaten me to it in 1239 01:31:12,200 --> 01:31:17,320 terms of moving over from CFD. So particularly Peter Forsyth, 1240 01:31:17,800 --> 01:31:21,640 University of Waterloo in Canada, who I think Peter had 1241 01:31:21,640 --> 01:31:26,520 come out of the oil reservoir CFD area if I remember 1242 01:31:26,520 --> 01:31:29,760 correctly. And so he he did a lot of the 1243 01:31:29,760 --> 01:31:35,920 pioneering work in terms of PDE methods, you know, numerical 1244 01:31:35,920 --> 01:31:40,520 methods for for finance. And so there wasn't, as it 1245 01:31:40,520 --> 01:31:45,000 happened so much, you know, left for, for me, as I had maybe 1246 01:31:45,000 --> 01:31:50,760 thought. And then I mean this, this was 1247 01:31:51,080 --> 01:31:53,240 over a relatively short period of time. 1248 01:31:53,680 --> 01:31:58,200 I thought that I would probably have to teach a course on Monte 1249 01:31:58,200 --> 01:32:01,880 Carlo methods, you know, because, you know, we, we, we 1250 01:32:01,880 --> 01:32:06,360 had still have an MSC in mathematical and computational 1251 01:32:06,360 --> 01:32:09,760 finance. And I, I teach numerics on that. 1252 01:32:10,040 --> 01:32:12,720 And you know, we would need to teach them about both sides. 1253 01:32:13,400 --> 01:32:20,040 And so I took a 2 day course put on by a couple of professors 1254 01:32:20,240 --> 01:32:26,120 from Columbia University in, in, in the US, put on in, in London 1255 01:32:26,120 --> 01:32:31,120 for, for London finance. People managed to, to, to 1256 01:32:31,120 --> 01:32:35,960 convince the, the, the department to, to pay my fees 1257 01:32:35,960 --> 01:32:39,040 for that. They, they, they, they gave me a 1258 01:32:39,040 --> 01:32:44,040 50% discount as an academic, but Even so, it was a costlier 1259 01:32:44,200 --> 01:32:45,960 course to attend. Anyway. 1260 01:32:46,560 --> 01:32:50,600 So I went to this course and they were teaching me about, 1261 01:32:50,720 --> 01:32:56,200 about, about Monte Carlo methods and talking about doing 1262 01:32:56,320 --> 01:33:01,520 sensitivity calculations. And I went up and talked to Paul 1263 01:33:01,520 --> 01:33:04,880 Glassman, the lead guy in in a coffee break and said, well, 1264 01:33:05,240 --> 01:33:09,360 this is all fascinating, but I presume that of course you, you 1265 01:33:09,400 --> 01:33:13,160 actually use an joint methods to to compute these sensitivities 1266 01:33:13,320 --> 01:33:17,520 more more efficiently. To which he said, what? 1267 01:33:19,920 --> 01:33:24,840 So yeah. So just pure, pure luck. 1268 01:33:24,840 --> 01:33:29,920 I was able to introduce adjoint methods to the financial Monte 1269 01:33:29,920 --> 01:33:34,400 Carlo community for doing sensitivity calculations. 1270 01:33:34,640 --> 01:33:41,280 So I did a paper with with, with Paul Glasserman in a finance, so 1271 01:33:41,360 --> 01:33:46,560 industry magazine really, rather than as a proper academic 1272 01:33:46,560 --> 01:33:48,400 journal. Because from my point of view 1273 01:33:48,400 --> 01:33:50,680 there was absolutely nothing new mathematically. 1274 01:33:50,680 --> 01:33:57,320 This was just a new application. This is a paper that went, went, 1275 01:33:57,320 --> 01:34:03,080 went by the title of smoking a joint, which helped helped. 1276 01:34:03,080 --> 01:34:07,200 It's it's notoriety. This is a journal or, or trade 1277 01:34:07,200 --> 01:34:10,360 journal that liked puns in their titles. 1278 01:34:10,600 --> 01:34:15,080 And so I would never dare do that in an academic journal. 1279 01:34:15,120 --> 01:34:19,680 But anyway, so yeah. So I so I got known for, you 1280 01:34:19,680 --> 01:34:23,880 know, the adjoint work, which is really sort of taken over in, in 1281 01:34:24,000 --> 01:34:28,960 in the finance sector. And then the multi level Monte 1282 01:34:28,960 --> 01:34:31,240 Carlo. Yes, I mean, you're right to 1283 01:34:31,240 --> 01:34:33,720 take, you know, the analogy to multigrid. 1284 01:34:33,720 --> 01:34:39,120 I mean, multigrid is such a fundamental part of CFD that it 1285 01:34:39,120 --> 01:34:43,400 was natural in getting into Monte Carlo methods to think, 1286 01:34:43,400 --> 01:34:46,960 well, is there anything analogous that that that we can 1287 01:34:46,960 --> 01:34:52,120 do here? And so, yeah, I came up with 1288 01:34:52,120 --> 01:34:54,480 with, you know, the multi level idea. 1289 01:34:54,680 --> 01:34:58,200 And it's one of those things that like multigrid itself, I 1290 01:34:58,200 --> 01:35:02,920 mean, it's such a simple idea. It really ought to have been 1291 01:35:02,920 --> 01:35:06,960 thought of ages before. But because I came in from a 1292 01:35:06,960 --> 01:35:10,240 different background, you know, this was, this was part of my 1293 01:35:10,240 --> 01:35:15,720 toolkit, you know, it, it, it was a fairly natural thing for, 1294 01:35:15,880 --> 01:35:19,120 for, for me to do. And so I've kind of been living 1295 01:35:19,120 --> 01:35:23,680 off that and extensions of that, you know, for the last 15 years. 1296 01:35:26,880 --> 01:35:30,200 And what about the computing side that the GPU side, you 1297 01:35:30,200 --> 01:35:34,120 know, is that is that being just because of, of, of an interest? 1298 01:35:34,120 --> 01:35:38,440 How much of that and forgive me for for not knowing, but how, 1299 01:35:39,080 --> 01:35:42,440 how much is that shaped Also on the maths side, how, how much is 1300 01:35:42,440 --> 01:35:45,160 that acceleration? You know GPUs to CFD is well 1301 01:35:45,160 --> 01:35:49,240 known but is it similar ID on the maths side? 1302 01:35:50,960 --> 01:35:55,680 So, I mean, most maths research just doesn't need lots of 1303 01:35:55,680 --> 01:35:59,200 compute power. I mean, these these days more 1304 01:35:59,200 --> 01:36:03,120 and more, you know, our, our students do things in Python 1305 01:36:03,440 --> 01:36:09,920 And, you know, maybe some will, will use the jacks package 1306 01:36:09,920 --> 01:36:11,800 within Python to get performance. 1307 01:36:12,120 --> 01:36:17,280 But a lot of work. No, nobody worries about, you 1308 01:36:17,280 --> 01:36:22,960 know, performance, you know, so I'm, I'm kind of an unusual, 1309 01:36:23,560 --> 01:36:27,200 yeah. So I, I, I still teach my, my 1310 01:36:27,200 --> 01:36:30,240 CUDA course every year with, with Wes. 1311 01:36:31,840 --> 01:36:35,160 You know, a few weeks ago I taught in, you know, a one day 1312 01:36:35,160 --> 01:36:40,760 open MP, you know, some mini course for, for PhD students 1313 01:36:40,760 --> 01:36:44,360 just to introduce them to this forgetting performance. 1314 01:36:44,560 --> 01:36:48,640 But there's very, very few students who are particularly 1315 01:36:48,640 --> 01:36:52,640 interested in, in that. It is an interesting question. 1316 01:36:52,640 --> 01:36:55,280 You know, where, where does that kind of work belong? 1317 01:36:55,280 --> 01:36:59,920 You know, to what extent does it belong in computer science or 1318 01:36:59,920 --> 01:37:09,080 maths or engineering? You know, I guess I'm somehow a 1319 01:37:09,080 --> 01:37:14,720 product of, of my time and it's not clear that there will be a 1320 01:37:14,720 --> 01:37:19,440 new generation of people like me, at least not in maths. 1321 01:37:19,680 --> 01:37:21,600 Yes. I'm not sure where, where the 1322 01:37:21,600 --> 01:37:26,680 next generation of me sort of lives, you know, so, so there's 1323 01:37:26,680 --> 01:37:29,200 Wes. Wes is in, in engineering now, 1324 01:37:29,640 --> 01:37:33,160 you know, having moved with, with, with OARC. 1325 01:37:33,760 --> 01:37:39,600 So, you know, Wes is now kind of my successor within the 1326 01:37:39,600 --> 01:37:47,600 university. He's, he's now Mr. HPC, you 1327 01:37:47,640 --> 01:37:53,320 know, there will be individuals in departments like physics and 1328 01:37:53,320 --> 01:37:58,280 chemistry and biochemistry who, who have their expertise in HPC, 1329 01:37:58,560 --> 01:38:02,200 but we don't really, you know, we're not a community as such 1330 01:38:02,720 --> 01:38:09,160 now, I would say, and we're still struggling a bit to 1331 01:38:10,000 --> 01:38:13,480 organise graduate teaching across the university. 1332 01:38:13,880 --> 01:38:19,680 I mean, this is something where Oxford and I think Cambridge 1333 01:38:19,680 --> 01:38:24,360 are, are poor compared to our American counterparts, where 1334 01:38:24,720 --> 01:38:27,160 you'll, you'll have graduate courses offered by one 1335 01:38:27,160 --> 01:38:30,880 department taken by people from across the university. 1336 01:38:31,520 --> 01:38:33,640 Yeah. We, we don't do enough of that 1337 01:38:34,160 --> 01:38:37,280 Our, our, our, our CUDA course is, is unusual. 1338 01:38:37,280 --> 01:38:43,520 I mean, this this year, I think we're currently up to about well 1339 01:38:43,520 --> 01:38:49,280 over 100 people signed up, of whom 65 are Oxford people and 1340 01:38:49,280 --> 01:38:56,040 another 35 externals, you know, and so the Oxford people do come 1341 01:38:56,320 --> 01:39:01,640 from across the university, but that, that's very unusual in, in 1342 01:39:01,640 --> 01:39:06,800 the Oxford setup. I'm, I'm, I'm trying to get more 1343 01:39:06,800 --> 01:39:09,520 of that happening. You know that we we have a much 1344 01:39:09,520 --> 01:39:17,240 more systematic training in, in advanced computing because you 1345 01:39:17,240 --> 01:39:19,160 know there are needs across the university. 1346 01:39:20,720 --> 01:39:23,800 I mean, I have to ask the obvious question, which is you 1347 01:39:23,800 --> 01:39:27,760 know, you've you've pivoted once I guess to the maps were you, 1348 01:39:28,040 --> 01:39:30,560 were you? And are you still tempted on the 1349 01:39:30,760 --> 01:39:34,680 AI given that it seems your skills are probably absolutely 1350 01:39:34,680 --> 01:39:38,160 perfectly aligned, which is strong maps under tech, 1351 01:39:38,160 --> 01:39:42,920 engineering and HPC? Yeah, So the quick answer is no. 1352 01:39:44,360 --> 01:39:49,120 So it, it, it is curious. I've I've got so three points of 1353 01:39:49,120 --> 01:39:54,760 contact with with AI. I've, I've got the, the GP US 1354 01:39:55,280 --> 01:40:02,560 I've got, you know, stochastic gradient method is very close to 1355 01:40:02,560 --> 01:40:04,280 the stuff I'm doing in Monte Carlo. 1356 01:40:05,640 --> 01:40:08,240 And there's the I joint, you know, so. 1357 01:40:09,920 --> 01:40:13,200 I mean, certainly at one point in the past, if you looked at 1358 01:40:13,200 --> 01:40:18,240 the source code for Pytorch, there were more references to my 1359 01:40:18,240 --> 01:40:23,680 adjoint publications than than any other academic, you know, So 1360 01:40:25,280 --> 01:40:28,360 it is, it is interesting how, how, how the adjoint stuffs got 1361 01:40:28,360 --> 01:40:35,960 got, got picked up. But, but no, no, it, it AAI is 1362 01:40:35,960 --> 01:40:39,640 for, for for a new generation. No, I'm, I'm, I'm not doing 1363 01:40:39,640 --> 01:40:44,960 another pivot. I'm, I'm I'm happy to keep 1364 01:40:44,960 --> 01:40:50,160 things going on on the HPC side. So I will continue doing things 1365 01:40:50,160 --> 01:40:55,120 with, with, with GP us I've, I'm, I'm interested in the 1366 01:40:55,120 --> 01:40:59,560 potential of Fpgas for doing finance calculations. 1367 01:41:01,520 --> 01:41:06,040 It it it's amusing when when I arrived in in 92 in computer 1368 01:41:06,040 --> 01:41:11,280 science, one of my colleagues then in in computer science told 1369 01:41:11,280 --> 01:41:15,280 me, you know, this is wonderful new technology Mike called Fpgas 1370 01:41:15,440 --> 01:41:18,080 is going to absolutely revolutionize everything that 1371 01:41:18,080 --> 01:41:19,720 you're doing. You know, you really need to 1372 01:41:19,720 --> 01:41:25,600 learn about it, and that's kind of still the message. 1373 01:41:25,840 --> 01:41:27,680 Yeah. I was just about to say they 1374 01:41:27,680 --> 01:41:29,960 haven't been right in the last 30 years. 1375 01:41:30,160 --> 01:41:32,600 Doesn't mean that they may not be right at some point in the 1376 01:41:32,600 --> 01:41:33,920 next 30 years. So. 1377 01:41:35,680 --> 01:41:37,400 And there's a story for GPUs from that. 1378 01:41:37,400 --> 01:41:39,760 It is that, you know, you were there early in the day, But I, I 1379 01:41:39,760 --> 01:41:43,200 guess it's Jensen always says that, you know, it's, it's taken 1380 01:41:43,200 --> 01:41:46,800 like 30 years to sort of get to this point where there's been 1381 01:41:46,800 --> 01:41:50,400 this massive, you know, use because of because of AI, but I 1382 01:41:50,400 --> 01:41:56,240 guess it was because of people who are early adopters it like 1383 01:41:56,240 --> 01:41:57,560 yourself. It's it's been a slow 1384 01:41:57,560 --> 01:42:00,640 progression, but suddenly it's all come together. 1385 01:42:00,640 --> 01:42:02,280 Hasn't. It, I mean, it has been 1386 01:42:02,280 --> 01:42:08,280 fascinating being sort of in, in, in the company of, of NVIDIA 1387 01:42:08,280 --> 01:42:13,640 during this whole evolution. You know, that, you know, I went 1388 01:42:14,160 --> 01:42:21,920 to all the GTCS in, in the early days and I, I remember being 1389 01:42:21,920 --> 01:42:29,040 there, you know, when Imagenet, you know, one, sorry, not, not 1390 01:42:29,040 --> 01:42:35,080 image net one, you know, one won the, you know, the competition, 1391 01:42:36,040 --> 01:42:40,480 yeah, for, you know, you know, for image recognition, image 1392 01:42:40,480 --> 01:42:44,160 clinic classification. And then the next year, the top 1393 01:42:44,160 --> 01:42:50,000 10 competitors won with, with GPUs, you know, and, and you 1394 01:42:50,000 --> 01:42:53,800 know, the way that NVIDIA pivoted was, was impressive, but 1395 01:42:53,800 --> 01:42:57,400 they were always on the lookout for the killer application. 1396 01:42:57,760 --> 01:43:02,280 You know, so right at the beginning they thought, at least 1397 01:43:02,280 --> 01:43:05,160 this is my impression that they thought that computational 1398 01:43:05,160 --> 01:43:07,360 finance might be the killer application. 1399 01:43:07,840 --> 01:43:11,400 And so that's why they were interested in in what I did in 1400 01:43:11,560 --> 01:43:14,400 implementing, you know, the first random number generator 1401 01:43:14,400 --> 01:43:20,120 on, on, on CUDA. You know, I think in, I think at 1402 01:43:20,120 --> 01:43:23,480 that point, the value proposition wasn't sufficient to 1403 01:43:23,480 --> 01:43:27,680 persuade the banks to take highly paid quants to, to 1404 01:43:27,680 --> 01:43:33,040 rewrite all of the software. And so the adoption rate was, 1405 01:43:33,040 --> 01:43:38,120 was, was slow. But yeah, yeah, the company was 1406 01:43:38,120 --> 01:43:40,600 always on the lookout for that killer application. 1407 01:43:41,000 --> 01:43:44,480 And, you know, when when they saw, you know, those early days 1408 01:43:44,480 --> 01:43:49,680 in AI, they thought, yeah, let's let's let's double down on this. 1409 01:43:49,680 --> 01:43:53,040 And, you know, first, first on the software side and then 1410 01:43:53,040 --> 01:43:55,040 increasingly on the hardware side as well. 1411 01:43:55,360 --> 01:43:57,320 Yeah, I mean, it is. Fascinating. 1412 01:43:58,200 --> 01:44:05,800 I just find that the CFD in some ways is still a niche industry 1413 01:44:05,800 --> 01:44:08,200 at a global level. You know the amount of money and 1414 01:44:08,200 --> 01:44:11,000 the amount of people. But I do always find it 1415 01:44:11,000 --> 01:44:16,360 fascinating that so many things have originated in CFD. 1416 01:44:16,440 --> 01:44:19,800 I mean, Ian Buck's PhD was CFD, right? 1417 01:44:19,800 --> 01:44:22,240 CUDA, as far as I'm aware, there's like a fluid dynamics 1418 01:44:23,440 --> 01:44:27,640 application, you know, like with yourself with the early days of 1419 01:44:27,640 --> 01:44:29,600 CFD. OK, now you've it's. 1420 01:44:29,600 --> 01:44:32,760 It's funny how I guess because it was one of the original hard 1421 01:44:32,760 --> 01:44:34,400 problems to solve. Yes. 1422 01:44:34,400 --> 01:44:42,720 I mean CCFDI guess was a prime driver for HPC for for a long 1423 01:44:42,720 --> 01:44:54,480 period, you know, So yeah, I mean, both in terms of as it 1424 01:44:54,480 --> 01:45:00,720 were open, open CFD and all, all, all of you know, the 1425 01:45:00,720 --> 01:45:04,400 nuclear weapons stuff as well. Yeah, which is sort of related 1426 01:45:04,400 --> 01:45:10,200 very closely, you know. So whereas now you know, within 1427 01:45:10,200 --> 01:45:15,520 the UK the biggest computers or the most IT spend is in the 1428 01:45:15,520 --> 01:45:20,280 banks, you know, you know, So you know, the amount of money 1429 01:45:20,280 --> 01:45:25,000 that Rolls Royce spends on compute per year, I would 1430 01:45:25,000 --> 01:45:30,040 imagine is, is less than any any one of the big banks in London. 1431 01:45:31,240 --> 01:45:33,440 I don't know that for a fact, but no. 1432 01:45:34,040 --> 01:45:37,160 Probably. That's the case, you know, of 1433 01:45:37,160 --> 01:45:41,120 sense. Yeah. 1434 01:45:41,360 --> 01:45:44,920 You know, so these days it it's the money in AI which is driving 1435 01:45:44,920 --> 01:45:47,520 the hardware development, you know, clearly. 1436 01:45:47,840 --> 01:45:53,960 And so, you know, you know, CFD is no longer that driver. 1437 01:45:57,120 --> 01:46:00,720 I guess CFD was also driving a lot of development of numerical 1438 01:46:00,720 --> 01:46:05,080 methods within academia, you know, and academia tends to be 1439 01:46:05,080 --> 01:46:08,240 at the sort of bleeding edge of the technology. 1440 01:46:08,320 --> 01:46:13,480 You know, again, all of those cheap bodies, Yeah. 1441 01:46:13,920 --> 01:46:18,520 Bright, bright young minds and well, seeing just how much they 1442 01:46:18,520 --> 01:46:21,320 they can squeeze out of this, this new hardware. 1443 01:46:25,560 --> 01:46:28,360 But yes, what are the big drivers these days other than 1444 01:46:28,360 --> 01:46:29,560 AI? Yeah, Yeah. 1445 01:46:29,560 --> 01:46:32,920 I mean that so much is is is based on that now. 1446 01:46:34,720 --> 01:46:38,840 Yeah, that's why I find it interesting that the and it's a 1447 01:46:38,840 --> 01:46:41,720 bit more of a controversial topic, which is the convergence 1448 01:46:42,240 --> 01:46:45,600 or the use of AI techniques for some of these disciplines. 1449 01:46:45,600 --> 01:46:53,720 You know, whether just as there's influences of using CFD 1450 01:46:53,720 --> 01:47:01,520 or numerical methods for AI, is AI got any use within CFD or or 1451 01:47:01,640 --> 01:47:05,960 you know, do you see it in the maths world that the actual use 1452 01:47:05,960 --> 01:47:11,000 of AI techniques or is it controversial or not sort of 1453 01:47:11,000 --> 01:47:12,960 proven? You know, just as you were 1454 01:47:12,960 --> 01:47:17,000 bringing in like techniques from CFDCAI applications in the 1455 01:47:17,000 --> 01:47:22,400 finance. So I mean, in terms of AAI 1456 01:47:22,400 --> 01:47:33,200 within CFD in general, I would say I'm a sceptic, but but to 1457 01:47:33,200 --> 01:47:36,200 some extent that may be that I just haven't spent enough time 1458 01:47:36,200 --> 01:47:39,080 to actually see what is is happening. 1459 01:47:40,120 --> 01:47:45,840 I think potentially the idea of an AI based turbulence model 1460 01:47:46,160 --> 01:47:49,960 might, might make sense. I mean, turbulence modelling is 1461 01:47:49,960 --> 01:47:54,360 such a challenging topic and there's so little progress I 1462 01:47:54,360 --> 01:48:01,120 think has been made in the last, you know, 25 years that AAI may 1463 01:48:01,120 --> 01:48:08,080 may be the answer there. I think in I think, you know, 1464 01:48:08,080 --> 01:48:11,640 for detailed CFDI think it it, it's going to need to continue 1465 01:48:11,640 --> 01:48:18,360 to be traditional CFD methods. You know that whether AI can 1466 01:48:18,360 --> 01:48:23,760 then get trained on the results of a large number of simulations 1467 01:48:24,080 --> 01:48:29,600 such that it does a pretty good proxy for the purposes of of 1468 01:48:29,600 --> 01:48:37,160 design optimization. Maybe I guess that doesn't 1469 01:48:37,160 --> 01:48:40,880 particularly interest or excite me. 1470 01:48:41,400 --> 01:48:46,240 But again, maybe I'm, I'm just showing my, my, my age. 1471 01:48:47,680 --> 01:48:57,520 I, I am interested in the impact of AI on research more 1472 01:48:57,520 --> 01:49:03,000 generally. So, you know, I'm really very 1473 01:49:03,000 --> 01:49:10,320 impressed by the capabilities of the latest ChatGPT, you know, 1474 01:49:10,400 --> 01:49:15,520 and you know, I think it really has improved hugely, you know, 1475 01:49:15,680 --> 01:49:18,480 over the last two years, say, you know, so the pace of 1476 01:49:18,480 --> 01:49:25,880 development is, is fascinating, you know, so, so I'm doing quite 1477 01:49:25,880 --> 01:49:31,200 a few experiments in different settings just just to understand 1478 01:49:31,200 --> 01:49:36,880 what it can do. So, so you know, I mean, just 1479 01:49:36,880 --> 01:49:44,000 yesterday I was getting it to, you know, tell me all about, you 1480 01:49:44,080 --> 01:49:50,400 know, Brownian motion, spatial white noise, space-time white 1481 01:49:50,400 --> 01:49:53,520 noise for, you know, for stochastic modelling. 1482 01:49:54,400 --> 01:49:58,840 And so put together a whole lot of information for me to, to, to 1483 01:49:58,840 --> 01:50:03,600 give to an MSC student who's doing a project with me, you 1484 01:50:03,600 --> 01:50:08,360 know, all all the way to it, you know, telling me about the 1485 01:50:08,360 --> 01:50:13,640 stochastic heat equation, providing me with some sample 1486 01:50:13,640 --> 01:50:19,400 code, implementing it in, in a simple, you know, finite 1487 01:50:19,400 --> 01:50:25,560 difference approximation. It, it tried to set it up to, to 1488 01:50:25,560 --> 01:50:28,240 use multi level. So understood about multi level 1489 01:50:28,520 --> 01:50:35,040 it it didn't quite get things right, but but it, you know, did 1490 01:50:35,040 --> 01:50:38,960 surprisingly well, you know, so I really wonder where we're 1491 01:50:38,960 --> 01:50:42,560 going to be in five or ten years time, you know, so. 1492 01:50:42,880 --> 01:50:48,360 Yeah. I think it's entirely possible. 1493 01:50:48,360 --> 01:50:52,760 I might even go as far as to say likely, that 10 years from now 1494 01:50:53,120 --> 01:50:59,720 you will have a proof assistant that is capable of looking at a 1495 01:50:59,720 --> 01:51:04,320 theorem and proof given the necessary background 1496 01:51:04,320 --> 01:51:09,200 information, and it won't be able to say with certainty that 1497 01:51:09,200 --> 01:51:16,920 the proof is correct. But it will on many occasions be 1498 01:51:16,920 --> 01:51:21,280 able to highlight bits in the proof that look dodgy or don't 1499 01:51:21,280 --> 01:51:29,280 like wrong. So yeah, So I, I, I kind of 1500 01:51:29,280 --> 01:51:33,400 think that AI, you know, you should absolutely never trust it 1501 01:51:33,400 --> 01:51:35,800 100%. You should always look at what 1502 01:51:35,800 --> 01:51:38,800 it produces. But the fact that it gets some 1503 01:51:38,800 --> 01:51:42,480 things wrong doesn't matter if it also gets some things right. 1504 01:51:42,880 --> 01:51:47,760 And it is, you know, it's very helpful. 1505 01:51:47,760 --> 01:51:54,120 I mean, another another thing I did with it a few days ago, it's 1506 01:51:54,120 --> 01:51:59,960 always irritated me with programming Intel CPUs that at 1507 01:51:59,960 --> 01:52:03,480 times their compiler is very poor at vectorization. 1508 01:52:03,760 --> 01:52:07,960 And so if you really want to get performance, you have to work 1509 01:52:07,960 --> 01:52:11,720 with the vector intrinsics, you know, the AVX 512 intrinsics. 1510 01:52:13,040 --> 01:52:18,200 And it's irritated me the fact that Intel doesn't provide you 1511 01:52:18,800 --> 01:52:23,680 with AC plus plus class that has operator overloading, you know, 1512 01:52:24,080 --> 01:52:26,720 to go along with all of those intrinsics. 1513 01:52:28,480 --> 01:52:32,200 And so I asked ChatGPT about this and it, it produced me 1514 01:52:32,200 --> 01:52:36,960 with, you know, it gave me AC plus plus class definition with, 1515 01:52:36,960 --> 01:52:40,000 with all of the operator overloading, you know, and I'm, 1516 01:52:40,200 --> 01:52:42,880 I'm not a good enough C++ programmer that I could have 1517 01:52:42,880 --> 01:52:45,800 done that myself. So just saved me, you know, 1518 01:52:46,680 --> 01:52:56,040 incredible amount of time. So things that in a sense are 1519 01:52:56,720 --> 01:53:01,160 are routine enough for a large group of people now become 1520 01:53:01,160 --> 01:53:05,800 automatically available to everyone else who isn't part of 1521 01:53:05,800 --> 01:53:11,080 that, you know, So, so as a productivity tool, you know, it 1522 01:53:11,080 --> 01:53:17,440 really is capable. So I'm, I'm doing this, I mean, 1523 01:53:17,440 --> 01:53:21,520 partly from, for my own benefit, partly also to just spread the 1524 01:53:21,520 --> 01:53:25,840 word amongst my colleagues, you know, to give them a range of 1525 01:53:25,840 --> 01:53:30,920 examples of, look, this is what AI can do for you. 1526 01:53:31,680 --> 01:53:35,640 Yeah. I mean, I agree. 1527 01:53:35,640 --> 01:53:40,360 It's, it's amazing what it, it's and and it's, it almost does 1528 01:53:40,360 --> 01:53:43,480 need people to evangelise or show it because if you haven't 1529 01:53:43,480 --> 01:53:45,680 seen it, you don't know. But once you see that, Oh my 1530 01:53:45,680 --> 01:53:47,760 God, this is. Yeah. 1531 01:53:47,760 --> 01:53:51,320 So I'm, I'm, I'm now kind of getting into that evangelism 1532 01:53:51,680 --> 01:53:56,960 group. Yes, Yeah. 1533 01:53:57,240 --> 01:54:02,560 It it it, it is fascinating. So here's maybe a question for 1534 01:54:02,560 --> 01:54:08,440 you as, as we get towards the end of this discussion, which 1535 01:54:08,440 --> 01:54:13,400 I'm sure we could carry on for hours, because I, I, yeah, 1536 01:54:13,400 --> 01:54:16,200 you've got so many interesting things that we didn't talk 1537 01:54:16,200 --> 01:54:18,000 about. But I'm I'm conscious of like 1538 01:54:18,000 --> 01:54:22,040 your, your time, but maybe more philosophical question, which is 1539 01:54:22,040 --> 01:54:27,360 if you were if you were now a undergrad. 1540 01:54:27,960 --> 01:54:31,680 Yeah. In today's world, what would be 1541 01:54:31,680 --> 01:54:38,640 your advice on a career trajectory, what to focus on? 1542 01:54:38,640 --> 01:54:41,240 Like, I know it's a very difficult question, but seeing 1543 01:54:41,240 --> 01:54:44,120 everything you've done, is there anything you would advise now, 1544 01:54:44,720 --> 01:54:48,920 somebody who is, yeah, 1819 or doing an undergraduate and is 1545 01:54:48,920 --> 01:54:54,040 thinking about, you know, academia, industry, Is it good 1546 01:54:54,040 --> 01:54:57,400 to work with industry to get that understanding? 1547 01:54:59,480 --> 01:55:11,920 Yeah. Yeah, it's tough because I mean, 1548 01:55:11,920 --> 01:55:14,920 I think the main advice is to try to do something that you 1549 01:55:14,920 --> 01:55:17,120 enjoy doing. You know, you know, you know, 1550 01:55:17,320 --> 01:55:20,280 you know, we don't want everyone to be doing the same things. 1551 01:55:20,280 --> 01:55:25,760 Everyone should kind of pursue, you know what, what, what they 1552 01:55:25,760 --> 01:55:31,640 like. If, if I mean if it was me again 1553 01:55:31,640 --> 01:55:37,960 now, you know, thinking of, you know, where I was as a teenager, 1554 01:55:39,400 --> 01:55:43,640 you know, I would, I would still probably be, be heading in into 1555 01:55:43,720 --> 01:55:50,120 mathematics. You know, I would. 1556 01:55:50,440 --> 01:55:53,440 I think by my nature I was always interested in 1557 01:55:53,440 --> 01:55:57,320 applications. So I think as a teenager I was 1558 01:55:57,320 --> 01:56:00,360 interested in maths because it helped me with applications in 1559 01:56:00,360 --> 01:56:03,600 physics that that was kind of the motivation. 1560 01:56:05,520 --> 01:56:09,640 I mean maths now one of the areas of growth I think is in 1561 01:56:09,640 --> 01:56:14,160 terms of mathematical modelling in medicine. 1562 01:56:14,800 --> 01:56:20,680 So, so I do think if, if, if I was maybe an undergraduate now 1563 01:56:20,800 --> 01:56:26,760 and, and thinking about where I wanted to head for PhD, you 1564 01:56:26,760 --> 01:56:32,080 know, computational methods applied in, in the sort of 1565 01:56:32,080 --> 01:56:37,360 medical area, I think may, may be an area right right now. 1566 01:56:37,360 --> 01:56:40,040 It's still challenging with the whole funding situation. 1567 01:56:40,040 --> 01:56:42,960 So, you know, we have difficulties linking up medical 1568 01:56:42,960 --> 01:56:47,120 departments with science departments and, you know, the 1569 01:56:47,120 --> 01:56:51,160 whole way the UK funding mechanisms work, but at least we 1570 01:56:51,160 --> 01:56:54,080 don't have the problems that, that, that the US has for us. 1571 01:56:57,240 --> 01:57:01,440 So yes, I mean that that's maybe what, what I would end up doing 1572 01:57:01,440 --> 01:57:05,320 if, you know, I was, I was 18 again. 1573 01:57:05,320 --> 01:57:09,240 Now I don't know. But I think the main thing is, 1574 01:57:09,360 --> 01:57:14,960 is to try to enjoy what it is you do. 1575 01:57:14,960 --> 01:57:23,160 And, and you know, don't don't worry too much about, about the 1576 01:57:23,160 --> 01:57:26,520 future, you know, you know, you know, when I started as, as an 1577 01:57:26,520 --> 01:57:28,400 undergraduate at Cambridge, I really didn't know what I was 1578 01:57:28,400 --> 01:57:31,520 going to be doing at the end. I think, I think now students 1579 01:57:31,520 --> 01:57:34,560 start as undergrads with much more of an idea of what they 1580 01:57:34,560 --> 01:57:38,120 want to do at the end of it. And maybe there's. 1581 01:57:38,200 --> 01:57:40,440 Pressure. Yeah, there is more pressure 1582 01:57:40,440 --> 01:57:43,360 now. I mean, back when I was an 1583 01:57:43,360 --> 01:57:48,240 undergraduate, so few people went to university that in a 1584 01:57:48,240 --> 01:57:53,520 sense you were, you felt, you know, guaranteed that you you 1585 01:57:53,520 --> 01:57:55,120 would get a good job at the end of it. 1586 01:57:55,120 --> 01:58:00,160 And people do do not feel that guarantee now, you know, so, so, 1587 01:58:00,160 --> 01:58:05,000 so I do, you know, I do understand and appreciate that, 1588 01:58:05,000 --> 01:58:12,920 you know, I certainly wouldn't discourage people from doing, 1589 01:58:13,280 --> 01:58:15,880 you know, you know, pursuing interests in programming. 1590 01:58:16,160 --> 01:58:18,960 I know, I know there's talk about AAI is going to do away 1591 01:58:18,960 --> 01:58:21,640 with all these programmers. No, I don't think so. 1592 01:58:21,640 --> 01:58:26,560 I think, you know, you, you, you, you still need people with 1593 01:58:27,280 --> 01:58:30,200 with programming skills. I think I would definitely 1594 01:58:30,200 --> 01:58:34,400 encourage people to develop their AI using skills, you know, 1595 01:58:34,520 --> 01:58:41,600 so and engage with these AI tools, you know, because they 1596 01:58:41,880 --> 01:58:48,120 they are hugely useful, you know, and yet at the same time, 1597 01:58:48,360 --> 01:58:51,880 you have to have the critical skills to look at what they 1598 01:58:51,880 --> 01:58:55,600 produce and ask whether it's right or not, you know, so, so 1599 01:58:55,600 --> 01:59:00,040 you absolutely still do need to have very strong understanding 1600 01:59:00,040 --> 01:59:04,240 of your technical area, but given that, you know, the AI 1601 01:59:04,240 --> 01:59:09,040 tools can, you know, make you more productive. 1602 01:59:10,760 --> 01:59:14,320 Would would you agree as well that I mean it's easy for both 1603 01:59:14,320 --> 01:59:21,520 of us to say this, but that the if you want to guarantee a job 1604 01:59:21,520 --> 01:59:26,960 for the next 60 years, having a maths and or engineering 1605 01:59:26,960 --> 01:59:30,360 background allows you to turn your hand to almost any 1606 01:59:30,360 --> 01:59:33,600 problems. And it's, it's quite a general 1607 01:59:34,400 --> 01:59:38,200 skill set that is desirable because you could turn to maths, 1608 01:59:38,200 --> 01:59:40,760 you could turn to CFD, you could turn to biology, you could turn, 1609 01:59:41,360 --> 01:59:44,120 you know, there's always a need for the sort of mathematical 1610 01:59:44,440 --> 01:59:51,120 simulation side of things. I mean, I, I, I think maths is 1611 01:59:51,120 --> 01:59:56,080 rightly viewed as, as a subject that has, you know, lots of real 1612 01:59:56,080 --> 01:59:59,400 world applications. I think, I think it's viewed 1613 01:59:59,400 --> 02:00:05,040 more so now than probably 30 years ago, you know, so you 1614 02:00:05,040 --> 02:00:07,960 know, you know, because of data science, because of AI. 1615 02:00:08,960 --> 02:00:11,960 You know, the importance of maths I think is much better 1616 02:00:12,320 --> 02:00:16,840 understood. We still at some time, you know, 1617 02:00:17,200 --> 02:00:20,720 have have trouble convincing politicians of, of, of that from 1618 02:00:20,720 --> 02:00:25,480 the point of view of funding. You know, we, we, we do feel a 1619 02:00:25,480 --> 02:00:29,400 bit hard done by, shall we say, in terms of supporting, you 1620 02:00:29,440 --> 02:00:33,280 know, you know, the underpinning maths that's so important for so 1621 02:00:33,280 --> 02:00:38,440 many applications. So yes, I think, you know, a 1622 02:00:38,440 --> 02:00:44,120 mass education gives you a very firm foundation for, for life. 1623 02:00:46,880 --> 02:00:50,240 But again, I would, I would emphasise doing things that you 1624 02:00:51,280 --> 02:00:55,040 enjoy, you know, you know, I think, you know, there are 1625 02:00:55,040 --> 02:01:01,280 people on, on the, you know, the creative side where I don't see 1626 02:01:01,280 --> 02:01:07,960 AI sort of taking over. I mean, whether it's creative in 1627 02:01:07,960 --> 02:01:13,160 the arts or being a chef or being a singer, you know, I 1628 02:01:13,400 --> 02:01:19,080 mean, yes, I mean, we'll see where, where, where AI gets to 1629 02:01:19,080 --> 02:01:23,280 in terms of, you know, generating your pop songs and 1630 02:01:23,280 --> 02:01:24,320 things. I don't know, yes. 1631 02:01:24,320 --> 02:01:29,560 I mean, maybe, maybe it'll turn out yeah, but people will, will, 1632 02:01:29,560 --> 02:01:31,640 will still want live performances and stuff. 1633 02:01:31,640 --> 02:01:33,400 No. So, so I think there's a lot of 1634 02:01:33,400 --> 02:01:39,000 stuff on the creative side which will will be important for the 1635 02:01:39,000 --> 02:01:43,400 future. But, you know, there's only so 1636 02:01:43,400 --> 02:01:48,120 much that AAI can do. You know, I, I think it, I think 1637 02:01:48,120 --> 02:01:51,680 it's best to think of it as a productivity tool that it's 1638 02:01:51,680 --> 02:01:55,120 important you engage with so that you you have those skills 1639 02:01:55,120 --> 02:02:01,120 to be productive. But don't, don't think that it's 1640 02:02:01,120 --> 02:02:04,600 going to suddenly eliminate huge numbers of jobs. 1641 02:02:06,000 --> 02:02:09,040 Yeah. Well, and you still you need at 1642 02:02:09,120 --> 02:02:13,040 least at the moment there's a huge market for people with HPC, 1643 02:02:13,560 --> 02:02:17,640 CFD, programming, maths backgrounds to develop these AI 1644 02:02:17,640 --> 02:02:20,960 models. So it's actually a a great time 1645 02:02:20,960 --> 02:02:23,640 in some ways to have those skill sets. 1646 02:02:23,880 --> 02:02:26,760 Yes, I'm not sure that we're, we're developing enough new 1647 02:02:26,760 --> 02:02:33,320 people in HPC, you know, and I'm not sure which degree programs 1648 02:02:33,320 --> 02:02:36,040 they're coming out of it. It's definitely not maths. 1649 02:02:36,040 --> 02:02:38,200 It's definitely not computer science. 1650 02:02:39,040 --> 02:02:41,320 Well, not, not Oxford computer science. 1651 02:02:41,560 --> 02:02:46,280 You know, places like Warwick maybe have a bit more or Bristol 1652 02:02:46,280 --> 02:02:49,280 maybe have a bit more of a focus on on HPC. 1653 02:02:50,120 --> 02:02:55,440 Now it's, it's a very actually good question because I guess by 1654 02:02:55,440 --> 02:02:58,640 definition, what is HPC is part of the problem. 1655 02:02:59,080 --> 02:03:01,680 And I, I would agree with you that. 1656 02:03:01,680 --> 02:03:07,080 And again, to link to the AI, it's even more important because 1657 02:03:07,640 --> 02:03:10,480 you know, as you know, big AI training clusters are 1658 02:03:11,480 --> 02:03:15,240 essentially what people would have called HPC classes before. 1659 02:03:15,240 --> 02:03:19,520 There is essentially no difference networking that you 1660 02:03:19,520 --> 02:03:22,080 Slurm or the OK now Kubernetes and things. 1661 02:03:22,080 --> 02:03:26,560 But but you're right, what course teaches? 1662 02:03:29,520 --> 02:03:32,040 Yeah. Yes, I mean, I know some of the 1663 02:03:32,040 --> 02:03:37,680 inside story on, on in Microsoft's development of large 1664 02:03:38,040 --> 02:03:43,000 GPU clusters and how they in, you know, involved a consultant 1665 02:03:43,000 --> 02:03:46,800 who is one of the world's leading HPC experts, you know, 1666 02:03:46,800 --> 02:03:50,720 so, so yes, absolutely. You know, doing, doing these 1667 02:03:50,720 --> 02:03:54,520 really large systems is is a massive HPC challenge. 1668 02:03:55,160 --> 02:03:58,040 It, it, it would be interesting to know, you know, the people 1669 02:03:58,040 --> 02:04:03,680 that NVIDIA hires as dev techs, what is their background? 1670 02:04:04,400 --> 02:04:10,080 You know, to what extent have they been formally trained in 1671 02:04:10,080 --> 02:04:14,240 HPC or to what extent is, has it been a passion? 1672 02:04:14,680 --> 02:04:19,840 And they've learnt on the job in various application areas and 1673 02:04:20,000 --> 02:04:23,920 they've, they've proved their skills and, you know, been hired 1674 02:04:23,920 --> 02:04:27,120 on that basis. That's that's a good question. 1675 02:04:27,120 --> 02:04:31,440 And I think maybe this goes not full circle, but half to our 1676 02:04:31,440 --> 02:04:36,640 discussion about with some of these OEOERC or these research 1677 02:04:36,640 --> 02:04:41,400 software engineers that at least what I see is that people get 1678 02:04:41,400 --> 02:04:45,800 their HPC skills often during their PhD when they're using a 1679 02:04:45,800 --> 02:04:50,200 HPC utility and they're sort of become best buddies with the 1680 02:04:50,200 --> 02:04:52,200 admin because they want to get higher up in the queue where 1681 02:04:52,200 --> 02:04:54,200 they have to sort of figure out some courses. 1682 02:04:54,680 --> 02:04:58,080 And so they haven't done necessary HPC course, but 1683 02:04:58,080 --> 02:05:00,760 they've had to do it to get access to the compute. 1684 02:05:02,040 --> 02:05:04,800 But somebody was there to manage the system. 1685 02:05:04,800 --> 02:05:07,000 So it's probably those like RSI think they call it. 1686 02:05:07,120 --> 02:05:08,840 Is it Rs ES now research software? 1687 02:05:08,840 --> 02:05:12,440 Engineers. Are sort of like this lifeblood 1688 02:05:12,640 --> 02:05:15,360 who support the people and are probably helping them. 1689 02:05:15,360 --> 02:05:22,880 That maybe is an underrated skill or need in the community. 1690 02:05:23,520 --> 02:05:29,040 But you're right, I'm I'm not sure of any HPC official 1691 02:05:29,040 --> 02:05:34,320 training, but maybe I need to do a bit of digging around to we 1692 02:05:34,320 --> 02:05:36,440 sort of take for granted stuff that we have. 1693 02:05:36,440 --> 02:05:39,000 We never realized that. What if all those people retire? 1694 02:05:39,000 --> 02:05:43,000 I suppose is your point, isn't? It like, yeah, yes, yes, I 1695 02:05:43,040 --> 02:05:46,600 wonder yes, where, where, where, where the next generation of 1696 02:05:46,600 --> 02:05:49,840 academics is, is, is coming from, but. 1697 02:05:50,360 --> 02:05:54,640 Yeah, yeah. But I, yeah, really appreciate 1698 02:05:54,640 --> 02:05:59,360 you talking. And I, I will put some links to 1699 02:05:59,480 --> 02:06:02,080 the the sort of episode notes because you, you know, I saw 1700 02:06:02,120 --> 02:06:04,640 you've got a great website as well where you link some of the, 1701 02:06:04,800 --> 02:06:08,280 you know, full list of your papers, some of the courses that 1702 02:06:08,280 --> 02:06:10,800 you're doing so that people could maybe read up because we 1703 02:06:10,800 --> 02:06:12,880 didn't get into all of your academic papers. 1704 02:06:12,880 --> 02:06:15,640 But I know you've done a good job of linking and some of the 1705 02:06:15,640 --> 02:06:17,240 presentations and things like that. 1706 02:06:17,240 --> 02:06:20,280 So I'll put it through. But yeah, I just want to say 1707 02:06:20,280 --> 02:06:23,320 thank you also collectively thank you because all the work 1708 02:06:23,320 --> 02:06:27,520 that you did in throughout your career has actually helped 1709 02:06:27,520 --> 02:06:30,880 people like me and others who work in industry to be able to 1710 02:06:30,880 --> 02:06:34,440 do CFD in an easy way that we take for granted now and don't 1711 02:06:34,440 --> 02:06:37,960 have to program 2G grids by hand like like you did. 1712 02:06:37,960 --> 02:06:41,640 So thank you from all the CFD people who take it for granted 1713 02:06:41,640 --> 02:06:44,040 now. And yeah, thanks for taking the 1714 02:06:44,040 --> 02:06:46,760 time to to speak to me. You're very welcome. 1715 02:06:47,600 --> 02:06:50,320 It's fun. Fun reminiscing about these 1716 02:06:50,880 --> 02:06:53,480 these things from days past. Yeah. 1717 02:06:54,360 --> 02:06:55,440 Great. Thanks very much.