做过四代TPU的AlphaChip团队创业了：AI设计芯片，芯片再反过来养AI

id="podcast-transcript" data-framer-component-type="RichTextContainer" style="transform:none">IntroductionAzalia Mirhoseini: Right now we can’t have much of co-design between chips and models because of this asymmetric design cycle for chips, because it takes so long, it takes so much time to design chips that the cycle is—there’s a mismatch between how fast we can create the next generation AI methods and how fast we can build the next generation chips. But if we can make our chips much faster, then we can enable this co-design and co-evolution of workloads, applications and chips altogether. Stephanie Zhan: This week on Training Data, we explore the biggest bottleneck holding back AI: compute and the design of chips themselves. Our guests are the founders of Ricursive Intelligence, Anna Goldie and Azalia Mirhoseini, the team behind Google’s pioneering AlphaChip project that helped design four generations of TPUs. They’re now applying AI to the entire chip design process, transforming the industry from fabless to designless. Anna and Azalia paint a picture of what a Cambrian explosion of custom silicon means. We’ll explore the holy grail topic of recursive self improvement, how AI-designed chips unlock radically creative new chip designs and why democratizing chip design could accelerate the path to superintelligence. Enjoy the show. Main ConversationSonya Huang: Thank you so much for joining us. We’re so excited to celebrate the announcement of Ricursive Intelligence, and congratulations on embarking on this new adventure. Azalia Mirhoseini: Thank you, thank you so much. Stephanie Zhan: To kick off, you’ve said that chip design is the compute bottleneck to advancing AI. Can you share and paint a picture of what’s happening now? What are all the bottlenecks to chip design today? Anna Goldie: I mean, I think when we were saying that, we were kind of alluding to this motivation that we had originally with our moonshot, which was this observation that neural networks, these concepts that have been around for decades, but the AI that came out of it wasn’t that effective until we had more powerful computer systems and chips. And we thought that now that we had very powerful AI systems, we could use those to kind of tackle the bottlenecks in chip design and, you know, create basically more effective compute. So if you see these, like, scaling laws, basically the more compute you apply to training a model or inferring with a model, like, the more intelligence you get out. And we’re seeing that there’s this mismatch. Like, we’re using—say GPUs are, like, originally designed for graphics processing, but we’re somehow repurposing them for crypto and then for training neural networks models—I guess they’re good at, like, large matrix multiplies. But if we could better co-optimize the models and the hardware, we could get more effective compute, and then we could push ourselves out on those effective scaling laws. Sonya Huang: You were the co-creators of AlphaChip at Google, which I think is used in four successive generations now of TPUs. Maybe take us back to that project. How did it get started? What were the key results you drove? Azalia Mirhoseini: Yeah, so we started the project in 2018, and we had some earlier version of placement work that we were interested in solving. And that was not about chip placement, it was about compiler and basically mapping neural networks to chips. So we did that project, we got great results, we published, and we were like, what are the highest impact projects that now we can kind of push this further and have real world impact? And that’s where the chip placement came in the picture. And we started the project working very closely with the TPU team at Google, because back then we didn’t know much at all about how chip design is, how the process is. So we worked very closely with them from very early on. It’s interesting because at the time AI was still very talked about, but nothing like—the scale was nowhere near what it is today. So we had to really kind of work through things and showing them data and, like, constantly iterating over our approach and hearing them what they needed from us, listening to them iterate through the process. And eventually we went from a research concept on chip placement all the way to something that was actually used in product and we could tape out and all that. Stephanie Zhan: I remember one of our very first meals together when it hit me that you were so customer obsessed and had so much empathy for the customer. And I asked you how and why, and you said, “Well, I’ve been working with the TPU team as our customer internally for so many years.” So I think that really kind of gave you that other perspective. What were the TPU team’s early reactions to what you had created? Anna Goldie: I mean they were, like, super skeptical. For example, like, Azalia and I, we were researchers. Like, we had read a bunch of research papers and we were like, okay, there’s this half perimeter wire length, that’s what academics report results on. So we made this RL agent that could optimize half perimeter wire lengths for a placement, and then we were kind of excited about sharing that with them. And we showed these results, and they were actually kind of angry at us, like, “Why are you showing us these results? We don’t care about half perimeter wire length. Like, we want routed wire length, like, horizontal and vertical congestion, timing violations, power consumption area.” And so, you know, we just kept listening to them and we were like, “Okay, so what is the thing that matters?” I think another part is just to make your customer feel like they’re part of it, too. Stephanie Zhan: Yes. Anna Goldie: So for example, we worked with them to create the cost functions that they cared about, approximations of those. So, like, Mustafa, who was on the TPU team, developed these very fast, like, congestion-cost functions, and then we could optimize—and density, and then wire length as well. And then we show results on those, and then we run the commercial tool and we show that this actually correlates to good results on the metrics that they do care about. Azalia Mirhoseini: Yeah. Sonya Huang: So I want to talk about floor planning. And I come from an EDA family, you know, and floor planning is the crown jewel of EDA. Can you say just for our listeners that don’t come from the chip design industry, what exactly that is, and how your technical methods help solve it? Azalia Mirhoseini: Yeah. So floor planning is a process where we place the components of a chip onto the silicon. And the complexity of this process comes from the large size of the input problems. Like, for a block of a chip—not the entire chip, like, a single block in a chip—the graph that we are optimizing could have millions of nodes, and all of these nodes need to be placed and routed, and we need to make sure that all these constraints that earlier we were talking about, like PPA, power area performance are optimized and all the physical constraints are met. So the placement problem has to do with how we can do this such that all those constraints given by these, like, very small technology node sizes given by TSMC and others are met. So we are dealing with a combinatorial optimization problem that’s large scale, but also all the metrics are hard to evaluate and measure. Sonya Huang: Got it. And this traditionally was done together with EDA software like a Cadence or a Synopsys, right? How does your method differ from the classical methods? Azalia Mirhoseini: So what we developed here was a learning-based approach to the problem, where we would train a reinforcement learning agent that would try out different ways to approach the placement problem, and it would learn through interactions with this environment that we built to learn from the positive placement and also the negative examples and iteratively improve itself. So one of the main differences between our approach and all prior approaches was this ability to learn from experience, which would enable the model to self improve. Just like a human expert that becomes better as they solve more instances of the problem and harder problems, our agent was able to kind of exhibit similar behavior. And that was a very major kind of delta and change between what we had in prior approaches. Sonya Huang: Yeah. And I remember with—I mean, if you take AlphaGo as an analogy—move 37 was kind of shocking because it was just so different from how a human player would move. Are you seeing something similar happen with floor planning and chip design? Anna Goldie: Yeah, we saw these very strange, like, curved placements. So there are donut shapes as well. I think humans would tend to make the macros very—so macros are memory components that are larger, and they make them very aligned and then they would put logic maybe in the middle and then the wires would connect all of these components. But if you make the shapes curved, you can reduce the wire lanes, which can reduce the power consumption and timing violations. But it’s just the complexity of making this curve placement was, like, beyond what humans would have wanted to take on or it felt risky to them. So yeah, these aliens. Sonya Huang: [laughs] Stephanie Zhan: What was the moment you realized that this had the potential to transform the entire end-to-end process, not just the floor planning stage, and when you decided that it had the potential to really be a company? Azalia Mirhoseini: There were a few, like, milestones. The first was showing that it works, that it gets the superhuman results on some blocks. And then later on the chip was actually taped out and it came back and it was actually working, which is a big kind of milestone, because it can be really, really costly if for some reason our AI missed something. So those were big moments, the real impact. And also on the algorithmic side, when we saw this kind of self-improvement properties and the fact that our models were becoming better as they solved more problems. Like, that is very—because when we get to that point, then there is no stopping for the AI. They can see more data points and solve more instances of the chip optimization problem than any single human can ever do. So that’s the moment that you’re like, “Yes!” So there’s a lot more potential here. But AlphaChip was targeting a module, a part of physical design. But chip design, as we know, there are many more stages to it and many more components. And right now we think that everything from the state of AI and the capability of AI, both on LLM side and other graph optimizations and other approaches, and the way that we can really scale up these algorithms and enable synthetic data, very large-scale distributed computing, everything is ready to tackle the end-to-end chip design optimization problem. So that’s why we’re excited about doing the company right now. Sonya Huang: Super cool. How do you think about getting the right training data for this market? I think it’s—you know, it is kind of the binding constraint to AI in many ways. And is this a market where you can have, you know, enough synthetic data or self play to be able to bootstrap? Anna Goldie: We’re excited about synthetic data. So obviously there’s some open source data, but it’s not that interesting or meaningful. And, you know, customers are willing to share data with us, but we don’t want to train our models on that because we want to keep their data private and siloed from each other. But synthetic data is where we think there’s the most promise. Like Azalia and I, we’ve been working on synthetic data approaches for LLMs and various code domain tasks across Claude and Gemini. And we see a big opportunity to get a scale of data that would go far beyond what any customer could ever share with us, like, orders and orders of magnitude more. Sonya Huang: With the TPU program, I guess at what moment did the chip experts stop doubting you? And as each successive generation came out, what was the progressive impact of what RL was able to drive in those designs? Anna Goldie: I mean, I think that our approach to the TPU team is like, every week we would show them data over and over again. Sonya Huang: Every week? Anna Goldie: Yeah, every week for like a couple years. Yeah. Because I mean, the stakes are very high here. Like, if there’s something wrong with this layout, like, the setup on the TPU team is like this. The TPU is quite large and complex, so they divide it up into dozens of blocks, and then each block is owned by a human or a team of humans. And that’s their responsibility to get this block right. If anything goes wrong, like, it’s their fault, right? And so they would generate their own layouts in collaboration with commercial tools, and then we would show them our layout, some AI-generated layout. It would look weird, it would be curved. And they would have to say, “Okay, like, I’m picking this AI generated layout over my own layout and I’m taking responsibility for that. I’m also suggesting maybe that, you know, some kind of incompetence maybe on my own part or something—” which is not the case, but—and then I think, yeah, it requires a lot of trust. We have to be better in every single metric for them to choose that. And we saw across each successive generation of TPU that we were being adopted in more and more of the chip block, in more of the area, and also that we were getting increasingly superhuman performance. We published an AlphaChip blog post, I think September last year, we showed this curve. Sonya Huang: Can you say a word on that, on the blog post and the superhuman performance? Anna Goldie: Yeah. I think just that across every single generation that we used, which I think there were three generations shown in the blog post, but it actually was used in another generation after that was published. Every single one, we were seeing more of the area and more, like, increasingly superhuman performance. Azalia Mirhoseini: Yeah. So basically the delta between AlphaChip layouts and the baseline layouts are also growing, which is like—it’s a property of AI and how it scales with data. Anna Goldie: It makes sense, right? AlphaChip was trained on more and more CPU blocks, so it gets better. Sonya Huang: Why is the company called Ricursive? Anna Goldie: Because we’re recursive. We’re AI for chip design and chip design for AI, so recursive self-improvement. In terms of, like, the name, the spelling of the name, so the name itself is Ricursive. So the initials RI, RIcursive Intelligence, are the first two letters of the company’s name. Sonya Huang: What does recursive self-improvement—like, why does that matter for the broader AI race? Just say a word on—I think this is a root node problem. Just say a word on it. Azalia Mirhoseini: Yeah. So chips are the fuel for AI, and scaling laws are driving much of the progress in AI, whether it’s on pre-training, post-training, test time and all that. And so the faster we can make chips that are more custom or better designed for the AIs that we run, the faster we enable this more efficient kind of compute. And that bends the curve for our scaling law, so that means we get to the next generation of AI faster and we can create and design better AIs faster. And so this is a type of recursive self-improvement that we want to see, because our AIs then can help our chips become better and we can design them faster and so on. So that is the recursive self-improvement loop that we are going after. Stephanie Zhan: One of the visions that you have is to transform the industry from just fabless to designless. What does that mean? Anna Goldie: So fabless was basically this concept of it used to be that people thought that no serious chipmaker could exist without their own fab. This was, like, obviously before, like, these multi-trillion dollar companies like Nvidia. You know, TSMC basically created this whole new world where we could have these incredibly valuable fabless companies. So we think that there’s an opportunity to create incredibly valuable companies that don’t need to have their own in-house design teams. So many companies, they serve these models at massive scale. They’re spending, like, you know, I think $100 billion-plus on AI inference alone this year, and it’s growing rapidly. So companies would benefit from, like, maybe custom chips to serve their models or train them, but that requires enormous teams of hundreds or thousands of human experts in house. And we don’t think that’s necessary, so we want to move towards designless paradigm. Stephanie Zhan: That’s fascinating. We’re seeing it come true with obviously Google and TPUs, Amazon with Trainium, also OpenAI and Broadcom, and maybe even Tesla. Sonya Huang: Do you think the future is then—even today there’s almost like model application co-design. Do you think there’s going to be chip application co-design, and even individual companies will have multiple chip architectures as a result? Azalia Mirhoseini: Yes, we think that we are going to see more and more co-design across the deep learning or AI stack, from models and data to software and all the way to the chips. And we are going to enable that. And co-design is really the secret or the path to more efficiency and more performance going forward. But right now we can’t have much of co-design between chips and models because of this asymmetric kind of design cycle for chips, because it takes so long, it takes so much time to design chips, there’s a mismatch between how fast we can create the next generation AI methods and how fast we can build the next generation chips. But if we can make our chips much faster, then we can enable this co-design and co-evolution of workloads, applications and chips altogether. Stephanie Zhan: One of the things that I loved that we talked about early on was that the value you bring isn’t just on reducing the cost it takes to actually design a chip or the speed at which you can accelerate the chip design process, though that is transformative in itself. It’s actually to unlock completely new potential applications and maybe custom silicon as well alongside it. Can you share a little bit about that? What is the Cambrian explosion that you might expect? Anna Goldie: I mean we think computing is going to be increasingly ubiquitous in every aspect of our lives. So, like, obviously there are things like AR/VR, there’s maybe chips in space, even like hearing aids. Like, there are experiences that aren’t possible unless you can serve them at, like, sufficiently low inference like latency or at low enough power. And we think that custom silicon can enable these applications. Azalia Mirhoseini: Yeah, we think that AI is going to be everywhere, in every kind of experience, every aspect of industry and life going forward. And there are chips that would enable these AIs to run. And given the scale, we would want them to run very efficiently and low power, high speed, all of that. And custom silicon is really going to enable that. And we want to enable custom silicon for basically any workload that is being run at sufficient scale. Stephanie Zhan: Since we have the ear of whoever’s listening on the podcast, who might some of your ideal customers be? Azalia Mirhoseini: There are a range of customers that we are envisioning. In the first phase of the company when we are building ways to accelerate, dramatically accelerate the chip design process, our customers would be chip designers like Nvidia, AMD, ARM, MediaTek and all of these companies that are already designing their chips. And all of them would want to pass their design cycle, and we can help them with this technology. But we don’t want to stop there. We want to enable any customer that has a workload or a family of workloads that they want to run that they’re running it at sufficient scale and they would benefit from custom silicon. We want to enable them to have that without having teams of hundreds to thousands of chip designers in their companies. I think that kind of goes back to this Cambrian explosion of chips that we can enable. Sonya Huang: I’d love to talk about the kind of existing incumbents in chip design. So, like, I grew up in a family of Cadence. Mom and dad both worked at Cadence, and I think chip design is a duopoly between Synopsys and Cadence. And I think they’re adding AI to their product suites. How do you see your company playing out versus the incumbents adding AI? Anna Goldie: I think that we see ourselves as sort of coming from the opposite direction. Like, we’re a frontier AI lab. We come from Google Brain or Anthropic, those kind of backgrounds, and we want to rethink or reimagine how chip design can be done. And we think that fundamentally in order to be able to—this isn’t like a point solution thing where we want to replace some module one at a time with AI. We want to reimagine and co-optimize different stages, and we think that an AI approach is necessary here. Sonya Huang: What do people from the chip design industry think of you guys? I would imagine there’s a range from excitement to extreme skepticism. Extreme excitement to extreme skepticism. Stephanie Zhan: Or extreme fear. Sonya Huang: Yeah, what do the chip design folks think of you all? Anna Goldie: I think a lot of them are excited to work with us as potential customers and we’re excited about them, too. And a lot of people are excited to come join us and work together. But yeah, of course, like, what we’re doing is very ambitious. So I’m sure many people are skeptical, too. Sonya Huang: Let’s talk about that for a little bit. Like, there was—you know, there was some internet uproar. You know, micro-niche internet communities love getting spicy, and there was some spiciness around AlphaChip and, like, people from media saying, like “Ah, it’s not really real for X, Y, Z reasons.” Like, what do you think was behind that? And what do you think was valid in that criticism? And what do you think is the important thing that they weren’t realizing? Azalia Mirhoseini: So whenever AI goes to a new field and does something disruptive, we would see reactions like that. And this is not necessarily just for in our case, it appears almost in every other. Anna Goldie: A bitter lesson. Azalia Mirhoseini: And we think yeah, the bitter lesson is part of it. Like, we are true believers in the bitter lesson. And it’s usually a little challenging to kind of like accept that a whole new technology or new way of looking into problems that people have worked on for decades is just coming and is solving things more end to end. And these people, like in this case Anna and I and the AlphaChip team, where we were coming from, not from an EDA kind of background, at least we had not worked on that problem space and now we could come up with solutions that were being productionized and all that. So there’s always this kind of a reaction in the beginning. But somehow the true kind of impact of our work was much bigger than just the problem that we solved. And I can mention some of those. For example, like, doing reinforcement learning, looking into this graph neural net optimizations were applied to many other problems across chip design, including the best paper award at DAC in 2023. Sonya Huang: Nice. Azalia Mirhoseini: The first author of that paper is now our teammate at Ricursive. And that was for physical design. We saw its adoption in other stages of chip design like synthesis and so on. And kind of like the work brought a lot of, like, attention to AI back in 2020 in the chip design field. And we think it’s like one of the most impactful things as a result of our project. Right now, there are conferences—like, we actually are involved in one of them, LLM-Aided Design that are just focused on AI approaches to chip design, which back in the days was not, not a thing. So there are so many positive things happened as a result that we are grateful for that. Anna Goldie: I had a funny interaction at ML for Systems workshop, which is another conference that we had started in 2018. And this person was like, “Oh, AlphaChip inspired my entire PhD thesis.” And I was like, “Oh, that’s wonderful.” And I was like,”So I’m just curious, like, how did you first come across the paper?” And he was like, “Oh, because of the controversy.” Very amusing. I think the thing that surprised me a bit was I thought that the people who would be upset by the work would be, like, physical designers whose jobs were at risk, right? And those physical designers were skeptical. It required a lot of data for them to be able to change their mind and accept that our method worked. But they actually weren’t the people that were upset. It was people who had developed prior methods in the field. And I think in retrospect that makes sense. There’s something painful about, like, you pour your time, your human ingenuity or soul a bit into these methods, and then people come from outside your field and then they’re using sort of in some sense simpler approaches. Like, they’re using things that scale with data and compute and you’re outperforming. It’s a little—it’s painful. Sonya Huang: Totally. Anna Goldie: But, you know, everyone, we can all build on top of this work. Sonya Huang: Yeah, we can all adapt. Anna Goldie: Yeah. Sonya Huang: Can I ask you the opposite question? So I love the bitter lessons answer. I then have the opposite question, which is that while the large language models became so good at coding, for example, why won’t they just naturally become good at chip design? Why does there need to be a specific chip design frontier lab? Azalia Mirhoseini: Yeah, so chip design has a lot of components, and some of them are language and related to language and code, but actually a big chunk of it is not related to language and code, and it has to do with these different properties and graph structure of the chip. And all these constraints that appear, like, these are really, really hard, large scale combinatorial optimization problems that required a custom way of—like, specific kind of approaches from an AI perspective. And our approach here is to apply the right method to every problem. And LLMs, we love them so much and we’re going to use them extensively. We are going to build LLMs that are really useful for some stages of the process, but they’re not going to be sufficient for all of this. So we are going to have our own AIs and specific optimizations for different stages. And those are really important, because unless we design them, those kind of modules that are really fast, really optimized, we can’t iterate around them with LLMs or with other methods. So our approach is going to be a very hybrid kind of multifaceted AI that is [inaudible] this. But we are true believers in AGI and bitter lessons, and we just want to be on the frontier of that with better chips. Stephanie Zhan: What do you think a world with AGI looks like? It’s so hard to imagine the world in ten years or even five, but what do you imagine for whenever that future is? Anna Goldie: I guess AGI, like what does that mean, human level for everything? Like, maybe you just have many employees that are just effectively a bunch of compute powering them. I think eventually maybe we can all take a vacation. Azalia Mirhoseini: It’s just going to be extreme cases of productivity for every single human being. And as a result, we can create a lot of economical value at a scale that is not possible today, and hopefully distributed at a scale that is not possible today. So I have a very optimistic view about the future—about the future when we have AGI. Just lots of good things are going to happen, and if you are careful about it, we can distribute the value to all humans, and that’s great. Sonya Huang: We’re gonna have data centers in space. Anna Goldie: [laughs] Yes, why not? Sonya Huang: And is that gonna require—actually, is that gonna require custom silicon? Because it’s like a different thermal footprint, right? Anna Goldie: Yeah, that’s great. Azalia Mirhoseini: Yeah. There are different requirements in space, from temperature, from all sort of like kind of resistance of the chip and so on. Anna Goldie: Yeah, even working on the Pixel phone, they have different corners. So, like, the phone has to be robust to different temperatures, like different voltage settings. And so yeah, you just have to increase in many more corners, I guess, for these chips. Sonya Huang: Yeah, that’s a good example actually of where the different properties of different—heterogeneous compute. You have different chips, different use cases. I think chips in space. Right now the whole data-centers-in-space field is structured around how do you make existing GPUs perform actually in space. But I think if everything for the [inaudible] for your company works, there’s going to be specific compute for space. Azalia Mirhoseini: Customization for space in this case. Stephanie Zhan: So you’ve already become incredible talent magnets in just the couple weeks since incorporating the company. We have Jiwoo, Yi-Chen, Dan, Ebrahim. What else are you looking for in terms of talent? Anna Goldie: I mean we’re definitely building out on the technical side. So every stage of chip design we’re looking for top talent, but also on the LLM side. Like, we’re a frontier AI lab, so we want people all the way from, like, pre-training, mid-training, post-training, RL training, you know, experts in evaluations, data. And also on the non-technical side we want, you know, operations specialists who can help unlock, like, recruiting, like chief of staff, that kind of thing. Sonya Huang: What do you hope to have accomplished in a year? Anna Goldie: We’re going to release our first product by a year, but yeah. Stephanie Zhan: [laughs] Anna Goldie: Strong commercial partnerships. Azalia Mirhoseini: Yeah, strong partnerships, and we want to have our first product in that timeline that we can offer broadly, not just to our partners, but more broadly to other chip designers. Stephanie Zhan: Anything you’re willing to share in terms of what that first product might look like? Azalia Mirhoseini: Yes, it’s going to accelerate the process. It’s going to tackle the long poles in chip design, and it’s going to be more end to end than the products that we have access to today. Stephanie Zhan: How do you see the role of a human engineer evolving in a world where the design process is automated? What did they end up doing instead with their time and talent? Anna Goldie: I mean, I think there’s, like—our company goes through various phases so, like, the answer to that question maybe changes over time. But, you know, one way to think about it is like, these engineers can come work with us and help us reimagine the process. So that’s one way. Stephanie Zhan: [laughs] Anna Goldie: The other thing is sort of like as humans start working with, like, Claude code or Cursor, you know, maybe they’re not doing as much hands-on coding, but they’re becoming amplified and becoming much more productive through these AI tools. So something like that. Like, for example, even just AlphaChip, we could generate these, like, [inaudible] optimal curves of, like, different trade offs. Like, every one of these points would have taken humans, like, weeks to generate, but we could generate many, many of them because it just takes so little time and compute for this model. So the human can explore, like, what is it we really want, what trade offs we care about? Sonya Huang: So we’re going to vibe code chips? That’s what I heard. Anna Goldie: Maybe that’s not quite—but yeah. Stephanie Zhan: So you’ve worked with some of the greatest of all time: Jeff Dean, Noam Shazeer, many, many more, very closely, Kwak Lee and others. What are some of the lessons that you’ve learned from them or experiences that have really shaped you, and how have they shaped your perspective today? Azalia Mirhoseini: So I think with all of these people that you—and they’re extraordinary. So just setting the bar really high. And all of them are also extremely passionate about what they’re doing. And that’s how Anna and I feel about this company. Like, we are so excited to solve this problem no matter what. So that’s something that we think is very important to actually be successful. And the other thing about them is they’re just really nice people. Anna Goldie: Yeah. Azalia Mirhoseini: And setting that nice culture, collaborative culture, where everyone is heard and everyone can do their best work, that’s, like, a very important, important thing that we want to follow their roots as well. Anna Goldie: That’s right. I think, like, Jeff, for example, like, incredible breadth, incredible depth, incredible speed, but also, like, very high integrity. Like, he treats every single person with, like, respect. It doesn’t matter if they’re, like, president of the universe versus, like, they’re an intern or something. Like, it’s equal treatment. And I think, like, we want to embody that, too. And I just remember, like, Jeff, he actually gave me, like, a performance review, and it was like, the corrective feedback or something. He was like, “Ask for more from other people.” And so I wanted to expect more from others, too. Stephanie Zhan: I like that. I like what you shared. I saw all these photos of Jeff at Neurops running with, like, many people, which I think perfectly encapsulates what you just said. He welcomes others in, and he motivates them and spends time with anyone, no matter what. Anna Goldie: I remember Kwok also. He was our manager almost for 10 years. He was telling us, deep learning makes all of us renaissance people. Like, we can make contributions in many different types of fields. And I think that’s true here as well. And he also said, like, have fun in this company. Actually enjoy. We think that’s important, too. We think if everyone’s having fun, then we’re going to do even better work together. Sonya Huang: I love that. Stephanie Zhan: All right, thank you so much, Anna and Azalia. We’re so excited for your first year at Ricursive, and we’re really, really excited about it being the frontier lab for AI chip design. Azalia Mirhoseini: Thank you so much. Anna Goldie: Thank you. <div class="framer-mcvrsd" id="