Hello and welcome to my latest edition of Silicon Grapevine. Today I've got a very special guest, Aart de Geus, who's executive chairman and founder of Synopsys. Aart, hello.
Hello, Nitin. Pleasure to talk to you.
Well, thank you for coming on. And I think this is a very special moment for you because you've been awarded the Robert Noyce Award from SIA for 2024. Tell me a little bit about that and how it feels.
Well, you know, it's always weird to get an award because you're so conscious that everything one has done has been surrounded by fantastic people. And so I understand why the award has to go to one person, but at the same time it's always for a team. But this award was special in a different way, partially because as of January 1st, I've transitioned the COO job to Sassine Ghazi, which is a fantastic choice. And at the same time, I have embraced a new quest. In the acceptance speech, aside of all the thank yous that are really heartfelt, I tried to explain a little bit what that quest was and why I feel like 16 years old again, having aspirations and thoughts for the future and finding something that's very meaningful. In a nutshell, the quest is I'm trying to understand the systemic complexity of the intersection between energy and climate. And just all of those words are big words, so we can go into it if you like at some point in time.
Yeah, no, we will touch on that a little bit later. But if anybody reads anything about you, they'll see lots and lots of accolades, you know, like the most important man in the silicon industry. You've been there since almost the beginning doing what you do. But that takes a lot of effort, I think, at the beginning. What inspired you to even consider electronics in the first place? Tell us, take us back to your childhood.
Well, if I look at the essence of the childhood, it's not been so much electronics as more the sheer joy of building stuff together with others. And both are equally important. I like the building, making something better, growing it, evolving it, and so on. But all the joy comes out of doing it together with friends. And maybe one kid characteristic, and you know how it is, at least the feeling was we used to have the infinite long summer vacations. And you would sit, I was always outside on the street sitting there with a bunch of friends and then you go around, so what are we going to do? I don't know, what are we going to do? Well, I don't know. And I would always be the person that out of nowhere would start the sentence, you know, we're going to... and I didn't know yet what was going to come out... build a tree cabin. And instantaneously all kids know what a tree cabin is, except they have all different thoughts about it, but they all know that's going to be exciting. And so then you run home and you steal some stuff from your dad's garage and you find some old piece of carpet and some plastic, and the whole summer project is now not any different than building Synopsys. You're on a mission. And in that process, the cool thing is whenever multiple people find something that they do together, and at the end you sit in this little cram thing in the tree and you're kings of the hill.
Well, yeah. Was there somebody in the family who was studying something like this, or were in the sort of electrical, electronics, mechanical, whatever?
You know, the only thing I remembered was sort of somebody we called uncle, but he was really slightly more distant, and he worked at the Swiss CERN, which was the center of nuclear research in Geneva. And for whatever reason, at some point in time he gave me a little box with three resistors, a capacitor, you know, clearly a reject from something. And as a seven-year-old, that was essentially the first connection. And then for whatever reason, I got as a birthday gift or so a little cardboard box which is how you can put together a radio without a battery and still hear something, quite amazing actually. And so just gluing it together, following the instructions, there was no insight on how it worked, but it was also an understanding if you didn't do it right, it would certainly not work. And so those are the touch points. Electronics things came all by accident. Yes, I studied electrical engineering, but there was another fantastic accidental piece of luck, which is once I finished studies in Switzerland, that's where I mostly grew up.
EPFL, yes, very famous institute.
Yes. And so I decided, well, you know, I'm not sure what to do, I'm going to go to somewhere else in the world and just continue to study by virtue of not knowing what to do. And so I happened to arrive at a university in Dallas. And literally the minute I walked into the secretary's office to sign up, some guy walked in who, you know, we talked a little bit and he changed the name of my advisor to his own name. I had no idea what an advisor was, so it didn't make any difference. He was the new chair of the electrical engineering department. I didn't know what a chair was, so that made no difference. And only years later I understood who this person was. His name is Ron Rohrer and he is the father of the program SPICE.
And so suddenly I was next to somebody who today would be viewed as one of the legends in the early years. And only gradually over years I got to understand the gravitas of that, the impact. But when it was time to decide on what the dissertation would be, the PhD dissertation, he had meanwhile already moved away from the university, was still affiliated as a professor, and said, hey, why don't you come a few days, stay at my wife and my place, and then we'll talk about this dissertation. And during that time I stayed there, I think three months, and he signed me up to join General Electric all at the same time because I could do the dissertation while working there. Turned out I did the dissertation at night, which may not have been the best idea, but it all worked out. But what is relevant in this is that in that process he connected me with all of his best friends, which turned out to be both the generation before him and parallel to him. And these were best-in-class people in the world. And maybe one of the most profound lessons that I now cherish and I've learned from that is even if you don't know where you're going to go, what your passion is and so on, the ability to work with people that are really good, that you trust, that you like, is a fantastic path. And unconsciously, by introducing me to his friends and the people that had been all these legends in hindsight of creating this field, it set the bar at a certain level, very high. And I think excellence is mostly determined by who you're surrounded with more than anything else. And that in hindsight was just a fantastic gift that is actually quite... I'm not sure if it's a US culture, but I think definitely Silicon Valley culture, where you just surround by people who are good at something and then you just get together and put something together.
I think that's a great observation because Silicon Valley is fundamental people that come from all over the world, so it's very broad and diverse is the word I was looking for. But it's also people that master simultaneously this quest for technology and for the economics of technology. And in the early 2000s, I coined the term for the semiconductor industry that everything we do is tech-economics. And I can go deeper into that if you like, but if you miss the technology, obviously you have no economics. But if you can't master the economics, obviously you're not going to have some new investment in new technology. And what Silicon Valley has been so great at is to constantly find that balance. And there is critical mass of these type of people here. And not surprisingly, around that there's a ring of venture capital that understood this and immediately say, oh well, here's the place to be if you want to define the future.
So you ended up in the US, you're surrounded by amazing people, best in class. And then I guess, I think it's all documented, your journey from there, so I won't go into that. But when GE and you finished the CAD division or computer-aided engineering division there, that's when you decided with two friends, two co-colleagues, to start Synopsys or Optimal Solutions. Tell us just the thought process there.
Well, you know, decisions too happen often by accident, by luck, or by circumstances. And the circumstance to understand is that GE at that time was in semiconductors but decided to get out of it. We were going to be laid off. And so the decision to do something was a necessity. But there are a couple key decisions. And the first one was that the two other members of my team was, hey, we will think about what can we do with the technology and do it in an ethical fashion, meaning we will not take or steal anything from GE, but rather see if we can seek their support. And through other set of lucky pieces and good moves, we were able to convince GE to let us go with the technology and some small investment. And ultimately, I think it turned out to be a great decision for GE because they got out altogether and they got ultimately a great return on some piece of Synopsys. But for us, it was great that it was with the support, the okay of GE that we did this, so unencumbered. And then started the journey of, well, now you build a company. And so it was again building something with people that were great to work with. And many trials, many learnings, and somehow the survival instinct brought us here. But also the luck of having exactly the right thing at the right time that ultimately had a massive impact on the history of digital design.
And Synopsys plays a big part in this. But what were, I mean, in the early days, I guess the challenges with any startup are sort of, okay, get customers, get revenue, get some ROI for the investors. What were the big challenges in the early days for you in doing that?
Well, for starters, we pretty quickly understood that we had actually a great asset, which was a really good prototype. And so it was not like, okay, here's some slides, we're going to revolutionize the world, you know, give us money. No, it was we have real results because GE was using the technology. And actually, literally in a matter of weeks, we were able to get the support of another company, Harris Semiconductor, that for whatever reasons knew what we were doing and adopted it. And even before we had full funding, sent us a check for $100,000. And this was like, well, for a venture capitalist, you're not talking just some idea, you're talking something that already somebody is using. And the somebody is using is so important because the speed of development is also very much a function of the speed of the feedback you get. And may sound strange, but in many ways the most painful of our customers still today are our best customers because you do something great, they say, oh, that's fantastic what you just did, and then seven minutes later it's like, but you know, could you not really? And you get sort of this constant drive that they have. Applies to ourselves. And so for our entire existence of Synopsys, I think we've always tried to be very aligned with the people that had the biggest ambitions, the biggest drive, and of course over time also the biggest economics.
So Synopsys, that was 1986, '87, something like that. And then for the last 38 years, I'm not going to ask you to document that, but what would you say were the big landmarks for you, milestones in that journey?
Yeah, maybe two or three of the big milestones. The first real milestone is when you have a customer buying for the second time. It sounds so small, but a lot of people buy some stuff and then that was that, that's the end of the story. And so a repeat customer is the first indication that you have something. And it's also part of this tech-economics loop, that wheel starting to churn. The notion of scaling the company to actually have customers that are not just the friends you already know, but very rapidly have a professional sales force. Pretty quickly we went to Europe and then to Asia. The payout was actually substantially fast, at least for that time. It's not a milestone, it's sort of more a landmark, the fact that at some point in time you go IPO. I don't think it's a milestone because you're essentially on a journey where you have only one goal, which is do things better for the customer for next year. And of course, part of that milestone sequence is driven by what was so well formulated and so acted at that time, which was called Moore's Law. And Moore's Law for me inculcated the essence of the psychology of our field. An exponential is actually a really demanding thing. No matter how you work today, you're already late for tomorrow. Exponentials are math functions that we're born unhappy, they always want more. But I love exponentials because it reminds me of the slide rule.
Well, the slide rule is essentially an exponential that's been pushed in a flat plane, so to speak. But seeing the thing is, it is always amazing how the human cannot really perceive the exponential because you see a little piece of linear behavior. But the fact that it's, you know, we look back at since Synopsys started, probably 10 million x of productivity, one cannot imagine that. One knows it as a very big number, but what it really meant for a field on a worldwide basis for digital design obviously is manifested by its impact. But underneath it's not one exponential. Moore's Law was sort of the center, it gave the thematic to our field, but it's many joining exponentials of big changes. And so another breakpoint for us that initially was not heralded as such was the fact that the fastest way of designing a design is by not designing it, but reusing it. And so IP reuse, which today we are by far the broadest distributor of, and ARM is larger, but they focused on the processor mostly, is that was a massive parallel path to the whole synthesis creating and reusing. Well, if you can bring those two together, that is productivity right there. And then around that, in more and more complexity, both going down all the way to TCAD, which is like atomic level simulation, but also going way up and participating and using all the AI capabilities in the very design. And you see my hands always going like this because I think silicon is essentially the foundation upon which now AI is being built. And for many years, silicon provided something and then you would write software on top of it. Now the AI has blossomed on its own and so it's becoming really demanding and say, hey, where is your next 100x? And that pressure is very legit because their tech-economics are greatly improving if you can do things 10 times faster.
Well, I think was it last SNUG or the SNUG before where you presented and you presented those exponential graphs on how AI is just exploding, the sort of design capabilities, and that just sort of brings it home, doesn't it?
Yeah, but you know, it's interesting because one should train oneself to say, yeah, AI exponentials are always exploding because the difference between going from 1,000 to 2,000 is an explosion if you're a thousand. If it's from 10 million to 20 million, it's the same explosion, same exponential. And so I love this notion that we have at least one math function that describes super fast evolutions. And of course those have existed in history. One of my favorite ones is the printing press, right? By around 1450, and by the year 1500, 20 million books, and that changed the world completely and utterly. And then of course the Industrial Age had the same characteristics with a few downsides such as the climate issue. But you see how big these things are.
Yeah, yeah. I mean, by the way, just in the milestones, I mean, one of mine was the ARC, building the ARC processor company with a team, and then Synopsys acquiring it at some point later, which is quite an interesting merger of our two paths.
Well, yes. And so there are many of these breakpoints. And actually, one should really align those less with the company and more with the industry, because then you have big impact. And I mentioned a couple of those in the acceptance speech. And one that was very fundamental, if you take '47, the transistor comes about. It is 1960 that simultaneously Kilby and Noyce invented the integrated circuit. And why is that so fundamental? Because suddenly things that do something, a transistor, get connected in the same medium, silicon or germanium at that time, and thus create systemic complexity at the next level. And then you can go another 15 years or so and suddenly it's CMOS that changes both the quality of the switching and the power that's not needed to keep it in a certain state. I think that right now we are entering a 20-year massive exponential potential, and that is the notion of multi-die, including the 3D and including unbelievably dense connectivity. Because if you can make between two chips hundreds of thousands of connectors per square millimeter, that's a lot of connections. That changes architecture completely.
