[MUSIC] >> Welcome to Category Visionaries, the show dedicated to exploring exciting visions for the future from the founders or in the front lines building it. In each episode, we'll speak with a visionary founder who's building a new category or reimagining an existing one. We'll learn about the problem they solve, how their technology works, and unpack their vision for the future. I'm your host, Brett Stapper, CEO of Frontlines Media. Now, let's dive right into today's episode. [MUSIC] >> Hey everyone and welcome back to Category Visionaries. Today, we're speaking with Brandon Lucia, CEO and co-founder of Efficient Computer, a chip startup that's raised 16 million in funding. Brandon, how are you doing today? >> Doing great, Brett, thanks for having me on the podcast. >> Now, I'm from super excited and I would love to just begin with a quick summary of who you are and a bit more about your background. >> Yeah, yeah, I'd love to share that. I'm Brandon, and I am CEO and co-founder at Efficient. I'm also been a professor at Carnegie Mellon University for about 10 years. I've got a long history going back 17 years to when I got my PhD at the University of Washington, it seems like history now. Then working on computer architecture and computer systems research, and recently, we took some of the research that we've been developing at Carnegie Mellon and we spun it up and turned it into Efficient. Got some really exciting things that we're doing at Efficient. >> Was this a master plan the whole time that you were eventually going to start a startup? Or where did that idea come from to spin this out and build a company? >> Well, my research for a long time has been focused on a variety of topics, but one of them that's been really consistent theme across all of the research that we've done and is really at the heart of Efficient, and what we're doing at Efficient Computer is physically constrained computer systems, and especially computer systems that are constrained by energy, the availability of energy, and how much you can do with a limited supply of energy. I've been looking at those problems on the research side for a long time, and my co-founders, Nathan Beckman, also a professor at Carnegie Mellon, and a long time ago now, VHD student, Graham Gobievsky, he's now our CTO, Nathan and I co-advised Graham, and we were working on, at the time of the very beginning, was sort of motivated by an ideological research question. We said, "What do you get if you do computer architecture, but you just erase the whiteboard and you start from scratch? You focus entirely, let's do energy efficiency more than anything else. So we're not going to do the incremental performance optimization, we're not going to squeak out the last 5% of computing speed for a big margin in incremental energy consumption. We're just going to focus on, let's do energy efficiency the best way that we can." So that was a long time ago, now seven and a half, eight years ago we started doing that, and we developed that research agenda over that period of time that was basically Graham's PhD years. We got to the end of it, and we just saw that we had something kind of really cool at the end. We saw something that was really amazing. It was substantially more energy efficient than essentially anything else out there. We had made kind of a new way of doing computing in the architecture that we ended up with. Around the same time we crossed paths with someone named Alex Hawkinson, he's our fourth co-founder and he's this serial entrepreneur, amazing at the commercial side of things, really visionary when it comes to seeing how tech can have an impact on the world and more broadly, and he helped us understand the commercial potential of what we were building. This was now a couple of years ago at this point, but we put all that together and we spun out the company and we started really developing an earnest round. What we're now building at efficient, which is the world's most energy efficient general purpose processor. That's what we're building, and it's the result of that research that we were doing at Carnegie Mellon. Was that scary to take the leap like that? It was because I didn't, I'll be honest with you, I didn't really know what I was doing at the time, didn't know what I was getting into. I'm the first time CEO, this is my first startup company that I've been involved with, and it was daunting, didn't know what we were getting into, everything from, you know, it's funny in the beginning, you have these questions like, okay, now what? We have to incorporate a corporation, right? Okay, so let's do that. How do we do that? Then of course, I was relatively naive. Luckily, I had a lot of good mentors along the way. Alex especially had done this before, and we got advice from really great advisors from the academics world that we had a professional network in already, and people that had done startups, and also the people that we met along the way. Something I picked up along the way in doing this, in developing efficient from the beginning, is that there's a lot of people that have done this one way or the other, and they see where you're at when you're starting out, and they want to help you if that pulse that they stepped into. And so there's a lot of people that are really ready there with advice, and with sort of experiential learning that they have, that they can share. So we've taken advantage of a lot of that, and that's been really helpful, kind of finding our way, and getting to the point where we are now. We're really on a great path in what we're doing in efficient developing, the energy efficient, general purpose process that we're developing in efficient. As you made that transition from academia to a startup life, what was the biggest challenge that you had to overcome? Yeah, there's many challenges. I guess you know this, you've probably talked to a lot of founders, and have a big list of all the things that can go wrong and can go right. One that I really had to do a lot of learning and a lot of reflection on was taking these relatively speaking, pure research ideas, things that there's a technical benefit. We can do measurements in the lab and we can see there's a clear technical advantage. We know that, right? And then you have to take that and package it and think about how do we go to market with these ideas? How do we make these accessible outside of the context of scientific literature where we see a clear technical advantage? And how do we translate that into more end-to-end benefit? And that's something that in academia that happens in different measures, and depending on who you ask, it's more or less important. Some people do it more than others. But in spinning the ideas out and developing a company around the ideas, we really had to take stock and we really had to understand not just, okay, we have this cool technical advantage. Let's go rule the world or whatever. That's not really how it works. You have to go out and figure out the market each. And we did a lot of learning. We did a lot of building around that. And the technical advantage that we have in just category defining energy efficiency in the processor that we've built, that is a starting point when it comes to the go-to-market strategy. And so we go out and we find places where that provides differentiating advantage in the market for customers so that they can do things they've never been able to do before because of the energy efficiency advantage that we provide. So it's a different way of thinking about the technical advantage that we have. Well, that was a big change if I compare to the currency of academia. It's not just about having the novel idea. It's about seeing the application and seeing the impact downstream for applications in the market. And if we look at your technology, maybe a good way to approach this, let's imagine it's Thanksgiving, you're sitting there at the table and maybe an older aunt or a grandmother says, Brandon, how does this work? Explain it to me in simple terms. How do you explain the technology in very, very simple terms? Yeah, it's easiest to explain the technology by reference to what the advantages that we provide relative to other things that are in the market today. So what we're developing at efficient and what we've built is a processor. And there's many processors out there in the world. And they take devices and allow them to be smart. So I would say to my grandmother around the Thanksgiving table or whatever. And in order to be smart, we need to use some amount of energy, think something attached to a battery. And it can consume signals about the world through sensors. And then it can do some processing on that data. And that takes some amount of energy to do all that processing and think about the world. And what we do is we make that amount of energy go down by 100 types. And so that difference comes from how we structure the way that that smart device thinks about the world. And now, so in a slightly more technical sense, the way that we describe this is we have a new way of representing a computation and of mapping computation from software into hardware. And that's really what we're doing is fundamentally a new what we call it into jargon is computer architecture. And this is a new way of representing computation and then carrying it out on a chip. So, you know, I don't know if we were Thanksgiving. I don't know if grandma would hang with me through that entire explanation. But that's really what we're doing. That's different is a new way of representing computation. And what we get as a result of that is one to two orders of mag that's a decrease in the amount of energy that it takes to do computation and to make devices in the physical world smart. When you're having commercial conversations, when do you see like an aha? Like when does this just light up for them and they say, OK, I get it. This is incredible. Yeah, that's a really cool moment when we get in the market. And we're talking to a customer and trying to understand what are your requirements? What are the applications that you care about? And then how does that align with what we can do? And so one of the things that we frequently will see is we'll start and we'll talk to someone and they'll sort of come into this like, yeah, yeah, I've seen low power before. And we help them understand that there's really a difference between thinking about low power and thinking about high efficiency. And what we're doing at efficient is really high efficiency. And that means you can do 10 to 100 times more with a fixed amount of energy rather than just doing whatever you're doing, having very little power while you do it. And so the aha moment for a lot of customers is when we talk about, OK, you're working on a battery powered device. Let's say you're doing infrastructure monitoring and you're going to put a sensor on every utility pole or every manhole or every meter of water pipe that runs through the ground or something like that. In that kind of application, you might think, well, if we want to have a long lifetime and a large deployment, we can't really do much with the device. But in fact, you can. And that's where it starts to really become the aha moment for people that live in that world of, say, infrastructure where we can say, instead of three months, how about five years of battery life for your device? And instead of collecting sensor readings once an hour, how about we collect sensor readings continuously? And we can do signal processing and we can do machine learning on those data as we collect them, essentially adding more features, adding more intelligence and adding more capabilities of the device. And at the same time, increasing the battery life by a substantial amount from weeks to months to years, depending on the application. And so when we start to get into that, where it's not just incremental 20% increase in battery life, which people have probably seen before, they've seen promises of low power. And instead, it's really this kind of major change in the battery life. And at the same time, we can bump the features that go inside the device. That's what gets really cool. Then people are seeing what we see when we think about the core technology I do, which is, it's really something fundamentally different that we're doing. This is a new category of computing device. We're doing computing in a really different way than the processors that are in the market today. And that allows us to get these big advantages in terms of energy efficiency. This show is brought to you by Frontlines Media, Podcast Production Studio that helps P2B founders launch, manage and grow their own podcast. Now, if you're a founder, you may be thinking, I don't have time to host a podcast. I've got a company to build. Well, that's exactly what we built our service to do. You show up and host and we handle literally everything else to set up a call to discuss launching your own podcast, visit frontlines.io/podcast. Now back to today's episode. Whenever you're doing something that's radically different and exponentially better, obviously requires a lot of education, what are you doing to educate the market and make sure that they understand that this even exists? Yeah, it's a challenge. And what we're doing really is different. So one of the things that we face when we go in the market is we say, what we're doing is really different. And people are sometimes scared of change because they think the change is going to mean a lot of impact to say their software development process or the way they integrate components into a built device. And the way that we do the education around this is something we're really proud of as a company and something that's really at the heart of everything that we're doing and efficient is we're a software company as much as we're a hardware company. So we're making chips. We're making chips that implement this new architecture and new computation extremely efficiently. But we bring this on a daily basis at the office. We need to have an exceptionally good developer experience. We need to make it possible for the average programmer to write code and to be able to deploy that code to our device. And so our software stack helps with this education effort because the software static lets people think about software the same way. You can write the code you've been writing forever. And that helps us to really hand hold customers through the process of understanding our architecture. You know, we're doing this new kind of architecture. Here's a little sidebar in the history of that. You know, the architecture that has dominated in the market for decades, 50, 70 years, something that we call a von Neumann architecture. That's a piece of jargon that means the way that computers have been for a very long time. And we're quite different. We're a post von Neumann architecture where we spatially lay out a computation across the chip. And that feels a bit foreign when you look at sort of the abstract diagram of what's inside of our chip. But when we show you the customer, when we show you that the software story looks just like it does today, things are much easier. You can think about doing software the same way you write programs in high-level languages. And that really guides people through the process of understanding, you know, what we're doing is different, but it still remains accessible. And it's really easy to use because of the software stack and the software story that we've developed around it. What about from a ICP perspective? How do you think about your ICP? Who is that ideal customer? Yeah, so the way that we think about these chips that we're making and the software that we're developing and the tools that we're developing, finding their way into the market, we look for a few things to find the ideal customer. The first thing we look for is an application that is, this is the perfect case for us, an application that is not possible today because of the operational cost of deploying a fleet of devices that carries out some tasks. Infrastructure is a great example of this, you know, imagine smart agriculture. I mentioned a few examples of infrastructure before. And there, you want to get as much value out of a deployment to devices as possible. So you want to have as many devices as possible, they can get higher precision sensor data. And if you can do more on the device, then you can have less technical infrastructure built up around your devices. So the problem with a lot of applications, the ones that we really gravitate toward as the ideal use cases are the ones that are really limited by energy. And so the consequence of that is we'll start doing analysis of an application. And we'll see a case where the operational cost, where you have like literally dude in a truck driving around, replacing batteries day in and day out because with a huge fleet of devices, there's always a battery that's dying. And so we can step in and we directly address the energy aspect of the equation. So if we're looking at smart agriculture application, if we're looking at smart infrastructure application, that's where we can really move the needle on that. And it's not just that we're improving applications that exist today. It's that we can take application concepts that are basically infeasible today and we make those possible. And that's what's really cool is identifying those kinds of application concepts that are just not possible today because they're limited by the operational expense associated with battery maintenance and the limitations of energy. We get rid of the limits of energy on those applications. We can allow those things to flourish and scale and really create the value of the world that we're after. That's the ones that are really exciting for us. Those are particularly exciting for us. If you think ahead to your 2024 priorities, what's keeping you up at night and what are those top priorities? Yeah, so where we are as a company right now is we are steamrolling a 120% full steam ahead getting to the tape deadline for our product's chip. And that happens later this year. We have an absolutely phenomenally amazing team. I can't say enough good things about the team. They're the strongest technical engineering team I've ever worked with. It's incredible. And I'm just like in awe to be working with this group of people. They're pushing toward our product tape out later this year. They're amazing. We're going to get there. There are so many balls in the air to deliver product tape out like this to get to a hardware tape out deadline to pull together the software that goes with it. There's a lot of technical dependencies that are all interlinked and fairly complex. And we'll stick the landing. But there's a lot to be thinking about all the time. And outside of the technical core of what we're doing, there's a lot of the strategic ecosystem building and then go to market strategy building. And we also need to be maintaining around that to continue to derive value out of what we're building to. The two of those things together, the sort of convergence of those two major strategic threats, technical and market strategy, it's challenging. And it's a lot to take on. And there's a lot of moving parts. And so like any startup, I mean, that's really what keeps me at the night is getting all those things lined up, keeping them on the rails, having faith of the team. We have an amazing team. I trust them. We're going to deliver on this. And it's really going to be amazing. But yes, it's a lot to look at it. I can imagine now. Talk a little bit about fundraising. So as I mentioned, I'm making sure you've raised 16 million today. What have you learned about fundraising throughout this journey? I've learned everything that I know about fundraising as a result of this journey, because I came into this quite I think it's been very interesting to understand, to the extent that I do now understand the priorities of institutional investors. And even going further back, the interest of angel investors and really getting involved in a startup, when it's really an idea and a song, that's fun to think about now. Going back and having those early interactions with the angel investors and then following that, doing an institutional raise, I think that one thing I've taken away from interacting, especially with Eclipse, they've been an amazing partner and they've been fantastic. And they led around, we've been learning a lot from them in terms of go to market, in terms of operating into complex technical organization. They have veterans from the industry, from the semi industry. They have veterans and developing go to market strategy, rather this kind of thing. So they've been awesome. They've been really helpful. I think in general, the best VCs out there certainly must be the ones that are there to help you. They're there to be partners, not the purely of financial engagement, but rather to be there with you and provide support, provide cover and provide experience that you can take and build more value in the company that you're building. From a go to market perspective, what would be the number one piece of advice that you can give to another founder? Oh, this is probably the oldest one in the book, but I would say talk to as many customers as you can, as early as possible. It's something I wish I could go to two years back in time and talk to more customers. It's always the knowledge that you get from customers, the nuggets of information that you start to see emerge at the consistent priorities in many customers. That's the kind of thing I wish I had more of it now. I wish I had more of it then. I think there's probably no company in the world that doesn't wish they understood their customers, understood the market better. And so, yeah, I wish early and often, and as much as possible, go be in the market and understand the priorities of customers. I think I believe that what we're doing is building a product that's really going to change the world when we bring it to market. And I think there's probably a lot of ways to launch a product that is absolutely the wrong product for a market. And I think the best way to do that would be to not talk to customers along the way. So, we've been putting a high priority on that and having as many of those customer conversations as possible. It's been great. It's been informing the feature set. It's been informing how we pull things together and the things that we prioritize in the design. So, I know that's the oldest one in the book, but honestly, that's just the most important thing. I think when it comes to the go-to-market side of things. And final question for you. Let's zoom out three to five years into the future. What's the big picture vision? Yeah. So, efficient right now. We're racing toward a tape-out deadline towards the end of this year. We're going to have product silicon landing at the beginning of next year. That'll be in small runs. We're really excited to be developing that into some proof-of-concept applications and going to scale. We're going to be driving to volume at the end of next year. And we're really excited to see chips at volume in the market really use cases. That's super exciting. Beyond that, we're looking at scaling not just volume and not just scaling the team, but also looking at scaling the architecture that we've developed itself. So, our first product, the one that I was just describing, that's focused on the sort of enterprise-class IoT infrastructure as a canonical example, sort of focused on maybe wearable applications, lighter-weight applications like that. Looking forward, we're aiming at higher performance applications. And that means we have to take the technology that we have developed, and we have a path we need to execute on this path to scale up in terms of computing performance. That's very exciting because it opens up a whole new landscape of potential applications where we need higher performance, but they also demand extreme efficiency. So, some really cool ones that we're very excited about, one in particular is in the aerospace sector. So, in small satellites where we want to do Earth observation processing, hyperspectral data, image data, looking at RF signal processing inside of satellites, there's a lot of really cool applications. And satellites are my favorite example of a highly constrained application. You never have enough energy, you have to cram everything inside of this tiny box, and if anything goes wrong, everything goes wrong. So, it's a really cupped-through environment. It's really tough to work in that environment, your resource constrained. We have a solution for future space systems, even with the V1 architecture, which is lower performance, but has extremely high efficiency, but especially when we scale the technology to this higher-performance V2 product point. It's very exciting to think about efficient silicon at the heart of a lot of these aerospace, especially small satellite applications going forward. And looking in the longer view, I mean, this is where we have these multiple products in the market, serving these markets. It's really cool to think of building the organization, growing the team. We are our people, and then growing the teams is something I'm really excited about doing over time as we get these products fueled into the market. Hazing, I love it. Alright, Brandon, we are up on time, so we're going to have to wrap here. Before we do, if there's any founders that are listening in that want to fall along with your journey, where should they go? Yeah, you can check us out at efficient.computer. That's the best place to find us. You can check out our architecture, you can check out what we're doing in detail, and there's lots of ways to hit us up, check us out there. So, please do efficient.computer. That's where we are on the internet. Amazing. Thanks so much for taking the time. Yeah, thanks a lot, Brad. This episode of Category Visionaries is brought to you by Frontline's Media, Silicon Valley's leading podcast production studio. If you're a V2B founder looking for help launching and growing your own podcast, visit frontlines.io/podcast. And for the latest episodes, search for Category Visionaries on your podcast platform of choice. Thanks for listening, and we'll catch you on the next episode. [MUSIC]
