The following interview is a conversation we had with Aron Kirschen, CEO of SEMRON, on our podcast Category Visionaries. You can view the full episode here: $7.3M Raised to Build Ultra-Efficient 3D AI Chips
Brett
Welcome to Category Visionaries, the show dedicated to exploring exciting visions for the future from the founders who are on 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 Front Lines Media. Now let’s dive right into today’s episode.
Brett
Hey, everyone. And welcome back to Category Visionaries. Today we’re speaking with Aaron Kirschen, CEO of SEMRON, an AI chip maker that’s raised $7.3 million in funding.
Aron Kirschen
Aaron, how are you doing? Fine. Thanks for the invitation, Brett.
Brett
No problem. Super excited to have you here.
Aron Kirschen
Let’s go ahead and jump right in.
Brett
Talk to us about what you’re doing there at SEMRON.
Aron Kirschen
So, in short, we make 3D AI chips. 3D means we pack a lot of compute on a very tiny silicon area, and we can do so by leveraging on a new semiconductor device that is ultra efficient so it doesn’t lead to overheating. If you peg a lot of compute into a very small volume.
Brett
How’d you get into this world? What’s the backstory behind the company?
Aron Kirschen
So basically, when I was the first semester here in Tristram, Germany, in the university I met in a kind of scholarship, I met my now Co-Founder and he was really into microelectronics, even in school, patented stuff or sense in that. And we discussed, actually we met in Cambridge on a study trip, whatever, and we discussed and discussed innovation and he told me about, okay, all this exciting semiconductor stuff. And actually I remember my first big memory with him together was in the Eagle Pub in Cambridge where we sat at a table eating the fish and chips thing and we sat at exactly the same table as Francis Crick, who invented or discovered the DNA structure or helix structure. So that’s what I remember.
Aron Kirschen
And we kept discussing through the years and ended up and having a good idea how to overcome the physical limits of current technology, but it was nothing that seemed to be realistic in like five years. We were more like playing around. But suddenly it made a lot of sense to actually patent this idea. And I remember then when we did and the university heard about it, they offered us to pay for that for the patents, but were like, no, even though we don’t have much money, we try to find enough family, friends, money to pay ourselves. And yeah, so the first thing was he wrote his diploma thesis on that and then he became my supervisor. So I wrote my diploma thesis on that as well. And I think that was more like the starting point before we founded Zemrun in 2020.
Brett
It’s an interesting time to found a company in 2020. If I remember right, that was deep in the middle or the early days of the pandemic. What was that like trying to start building this company in the midst of the pandemic?
Aron Kirschen
Yeah, so the biggest impact of that was that we filed or we tried to get a visa to move to US to start the company there. But the embassy was like, okay, no, we are not even. Yeah. Looking at new applications right now because it’s not going to work anyway. So it was in, I think summer 2020 or autumn 2020 and were, yeah, we participated in the Berkeley Skydec program, which is usually a program in presence and in California and in Berkeley. And then of course it would have store right away founded in us but because it wasn’t possible. We are a German company right now, so good for Germany in that way, I would say we had to find a legal structure where we could put our patents for the investors and not have it privately.
Aron Kirschen
So that’s why we had to found the German GmbH, which is definitely not the best structure in the world to have a startup.
Brett
How did it feel for you in I think it’s November 2022 and ChatGPT really entered the mainstream and started to be talked about very widely. What did that moment feel like for you? Did you think, oh boy, like things are about to get crazy now or was it just, you know, another day in the office, the next day?
Aron Kirschen
So actually in November, to be honest, it became a big thing in December for us. But I also have to say we had a great advisor in Berkeley and he’s really into AI and he was one year ago before that happens, he was like, okay guys, you have to have a look on transformer based large language models. He didn’t use these words. I remember we didn’t have the wording, but were keeping an eye on it and tried to make sure that our hardware is fit for that. So when it happened it was like, great for us. So we are kind of prepared. But the impact on society is something extremely hard to foresee for those who are not really into the bubble. So in December were first trying that and I think in January 2023 it became a big topic everywhere.
Aron Kirschen
Not just in our office, on the streets, you know, everywhere.
Brett
If we zoom out from your perspective, you know, how do you think about the state of Edge AI hardware today?
Aron Kirschen
That’s a good question. So you mentioned ChatGPT. So it was probably for non tech people the ChatGPT moment where like was like. I think it will be a historical moment. I think it’s not a bold thesis but everybody was expecting okay now with ChatGPT everything we will live in a different world tomorrow and kind of we did. When it comes to edge devices, we still miss this kind of chatgpt moment. Like chatgpt makes a lot of sense in data centers and of course if you use it to Google smarter or make your research smarter and write messages smarter, it can help a lot. But it’s not really a game changer I would say in for smartphones or wearables in general. So I think we are still missing the color app for these kind of devices.
Aron Kirschen
Even though we have a lot of AI functionality in smartphones today. I think a lot of new things will come and we believe, and of course we might be a little bit biased here. We believe the reason why we don’t have the killer app is because you don’t have the hardware. And with not having hardware, I mean we won’t have it with the current approach in five years from now because there’s no way current technology can do that. To run for example a large language model for live translation like continuously in a discussion there’s nothing you can do, right? Or nothing without battery is down in like some minutes. And if you have this functionality, this could be a paradigm shift and could create a completely new devices similar to what happened when we got smartphones and not just phones. Right.
Brett
How do you think about the location that you’re in? I know in the pre interview were talking there a little bit. That’s a major hub there in Germany for those of us listening in. And you know, keep in mind a lot of our audience is here in Silicon Valley. What’s the area like? What’s going on there? Tell us a bit more.
Aron Kirschen
So in general I love Germany. I love the culture, I love many things about Germany. The food is great, it’s extremely underrated. But I think there has been better times for startups in Germany. The economy is not doing very well and for startups, I have to admit I’m not sure whether I would found again in Germany. So it’s a tough place when it comes to bureaucracy. I think that everybody knows that already. But I think in Germany it’s probably even better than other places in Europe. I lived for half a year in France. So there, when it comes to bureaucracy, it’s kind of probably even worse in some fields.
Aron Kirschen
But the thing that is that worries me most in Germany is the level of education or the decrease of the quality and the mindset of the young people, like in my age, like 25 to 30 people that are talking primarily about work, life balance and not about the exciting stuff they could do in their life when they are young and they have a lot of energy. So that’s something that worries me. And you have to create your own bubble. You can still do that. You can have very great people around you in Germany. It’s still a big country, but it’s not as easy as I would assume it’s in other places in the world.
Brett
And what’s the history there of Dresden? I understand it’s kind of the center there of microelectronics. You know, what is that history? What’s that backstory?
Aron Kirschen
Exactly. It’s kind of an interesting story because it dates back to the socialist time of this part of Germany. Right. We had a huge program in the GDR, the Democratic Republic of Germany, where they really wanted to do some semiconductor stuff for some time. It seemed like they would succeed with that. That was the first Megabit chip and so on. But of course it was still socialism, right? But when the curtain or Berlin Wall was falling, opening then we still had a lot of talent here and a lot of industry around that center. And so it was kind of obvious for the new leaders, mostly from western Germany, to try to concentrate or put a lot of money into the existing structures. And that’s why Germany became the center first of Germany for all the semiconductor stuff.
Aron Kirschen
And now it’s even center of microelectronics on a European level. So it’s definitely not Taiwan, but we have soon TSMC here, we have Infineon, we have Global Foundries, we have Bosch. There’s a huge fab here, xfab. So we have a lot of fabs here, a lot of suppliers. Definitely bigger than everywhere else in Europe. So from that perspective, if you have a semiconductor startup, Dresden is a great place.
Brett
Are you seeing a lot of other startups there? Is there a lot of innovation there? What’s going on there? You know, just from a startup perspective.
Aron Kirschen
Yeah. It’s not like Munich. You have a huge tech scene, a tech scene here. But I think the percentage of the hard tech companies in Dresden is definitely higher than any other city. I don’t have statistics at hand, but it’s what Is my experience. Usually you have Munich and you have Dresden and you have Berlin for all the software staff, you have Frankfurt for the fintech. So within Germany it’s definitely a good place for hard tech companies, sensor companies, other semiconductor startups. Yeah, so we have a small but nice group of startups. Interesting.
Brett
Let’s imagine that I’m visiting you there in Dresden or walking around a holiday market drinking some Glue vine together. How do you explain your technology and what it does in simple terms.
Aron Kirschen
So first of all that’s very important to mention here, I guess is we have the oldest Christmas market in the world. So it’s 590th Christmas market we have this year. So I think one of the most famous ones that’s very important to mention, I guess apart from that, how would I explain it to somebody on the Christmas market after three Glue Vines, I would say, look, when we calculate or when we carry out calculations today, it’s after all based on electrons moving around creating a lot of noise.
Brett
Right.
Aron Kirschen
So it’s somehow how we calculate. And noise is a bad thing because it creates heat and heat is energy consumption after all. And we need to overcome that to some point. And what if I tell you, not going into details necessarily here, but what if I tell you that we can also do the calculations by avoiding electrons moving around with a so called capacitive approach, which basically means you have something like a capacitor with an insulator in between where you carry out your operations. Not like in a transistor or something like that. It’s very simplified and it’s still using some technical terms, but you can think of that as a new transistor, the vast fundamental technology for GPUs and CPUs, but it’s highly specialized on operations we need for AI, so you can’t replace transistors with that device. It’s really an AI inference technology.
Brett
I see. Okay, that makes a lot of sense. Let’s see if you can do that with 10 blue vines. We’ll test it Next Christmas.
Aron Kirschen
Yes, sounds good. Perfect.
Brett
Let’s talk about go to market. And I understand with hardware it’s a different game when it comes to go to market. How are you thinking about the go to market strategy? How are you thinking about bringing this technology to market?
Aron Kirschen
So yeah, as you mentioned, we are. As a semiconductor startup, you don’t have the timeline or the roadmap usually of 1 year for go to market to really have to think in three, four, even five years. It depends on where you start. If you start with an existing semiconductor technology and just like have to design the chip, it’s still very complicated. But you can think in one, two years. If you start like us with a new semiconductor device, you really have to open the semiconductor process and make some changes there together with the foundry. Then you talk about three, four, five years. And if you really talk about new materials other than standard silicon or semiconductor CMOS technology, then it’s really complicated. We might even talk about like eight, nine, ten years. So that’s our horizon in semiconductor startups when it comes to Xamarin.
Aron Kirschen
I mean, we are four years around now and we are close to process freeze. That means we kind of developed the device and fabricated that and tested it and now can really start thinking about building demonstrators. And for us, it all comes down to these demonstrators going out to customers, showing them our FPGA together with our hardware. And what is probably also important to mention, we do not or will not likely won’t have millions of customers. For us, it’s more like having five, six big entities that is generating millions of revenue, that also shapes how we interact with these customers. It’s very intense, very technical, very detailed, of course, but it is less standard marketing. It’s not that much based on emotions, convincing people to use it. I would say it could be different, but in our case it’s really easy.
Aron Kirschen
You have to sit down with their engineering team, with our engineers, and figure it out how we can make it work. And that’s the stage we are at now.
Brett
Are there any semiconductor companies, you know, old ones, the previous generation, that have inspired your approach or just inspired how you think about building a company?
Aron Kirschen
There have been a lot of companies that inspired me. I not sure whether it’s necessarily from semiconductor space. Of course you have all the Apple stories and everybody’s familiar with that. But it’s not that interesting here because for probably the audience, because I still admire what Elon Musk did, it’s basics, for example. So these are the standard answers, I guess. I know I started to Read a little bit about the early days of Netflix and their startup culture seems to be very similar to ours. So usually us companies, not necessarily hardware.
Aron Kirschen
The more I get into these networks in semiconductor world, I also realize, okay, there have been a lot of companies in the 80s 90s that shaped the world and today not really known anymore, but we have a lot of people now in the 70s and 80s that were probably like us, very ambitious and with a great story. And I realized at some point most of the great people and advisors kind of can be traced back to National Semiconductors, which I don’t know. Do you know the National Semiconductors?
Brett
No, I don’t.
Aron Kirschen
Yeah. So that’s kind of interesting because I mean, today everybody knows Nvidia, right? And probably National Semi was never like the most valuable company in the world, but still it was probably a huge impact and it’s completely forgotten. I mean, it’s not hard to foresee that Nvidia, in 30 years for people that are less than 30 years old could be something like, I’ve heard about it, but I’m not really sure. Was it company, right? It could be. Right. And that’s kind of interesting because at some point when you realize, okay, there’s something beyond your horizon and there’s kind of a continuum and you just started at some point to catch up. So let’s see. So back to your question.
Aron Kirschen
I would not be able to name you a European semiconductor startup that I have that I try to learn from them because I think they did an awesome job.
Brett
That’s totally fair.
Aron Kirschen
That’s totally fair.
Brett
Now we’re somehow almost into 2025, so that’s why we’re talking about Christmas markets. We’re recording this five days before Christmas as you think ahead to 2025. What’s the big plan? What are some of those big objectives that you have?
Aron Kirschen
So actually it’s about customer engagement. Now it’s the first time for us to really sit down together with the engineers. And for us, even though we are hardware company, so today hardware means you have to do a lot of software. And this is kind of nice in some way because what we do right now is we get proprietary AI models, really exciting benchmarks or model architectures. In general. And we have to deploy that on our hardware that we currently have to emulate and we have some test structures and so on, and then we try to make it fit into that. And so we discussing with them about, okay, what can we do about your models? And come back with a nice performance simulation. And that is really based on our hardware aspects.
Aron Kirschen
And this is something that, for a semiconductor startup that for three, four years worked like it feels, it felt like swimming in the ocean. And you never had the real feedback, like mostly the market feedback you get from very superficial parks with them. Yeah, we might be interested, blah, blah, and probably investors that talk to our customers or introduce it to them just to get a sense whether it holds potential or not. But now it’s really like we understand what they need in an application. And of course it completely reshapes your coordinate system. You never thought about that before. And in next year I know that we will have so many discussions with actually very big, well known companies that are market leaders in their application or in that vertical that, yeah, it’s going to be a very exciting time.
Aron Kirschen
I’m really looking forward to next year. I can’t wait to start. I mean, I’m looking forward to Christmas as well, but I’m really excited that 2025 will start.
Brett
Yeah, it seems like 2025 is going to be the big year for just tech in general. So everyone I talk to, just myself included, a lot of excitement. So hopefully 2025 is awesome. Moving into our final couple of questions here, let’s talk a little bit about money. So as I mentioned there in the intro, 7.3 million euros raised to date. What have you learned about fundraising throughout this journey?
Aron Kirschen
I think my biggest learning was that I tended to be too arrogant in the beginning because, you know, you spend in fundraising, you spend like talking to 200 investors, I don’t know, giving all the team the same pitch. And of course, when you learn they didn’t even get the most simple thing, then it’s easy to say, okay, so yeah, it’s those stupid people, right? What I learned is that it might be true for some of the investors, but most of them are pretty smart and there’s a small fraction of them that I really admire. I talked to analysts, mostly from US VCs in the bay Area, not asking for money, but mostly asking for advice. Okay, what would you consider as a milestone to consider as dependable in our CSA after this round?
Aron Kirschen
And it’s just to get a sense of that and they even have on the Analyst level. They had people, it was just incredible how fast they were able to grasp what we are telling them. So this was an extremely incredible experience. But I still think also in Europe, many investors, they really are in a good position to get a big picture, to learn from different sources, obviously. And even though they don’t tell you, which is probably a problem of our culture in the western world or at least today, but they know a lot of things and asymmetric information. And that’s something probably my biggest learning.
Brett
From that final question for you. Let’s zoom out. I don’t know how far you want to go. We can go three years into the future, we can go five years into the future, we can Go 10 years into the future, however far you want to go. And what does the big picture vision look like?
Aron Kirschen
So especially for someone, you asked me, okay, what’s going to look like for someone in three or five years? Yes, okay. So yeah, that’s kind of easy. I mean, obviously I had that vision. You know, every revolution became the impact on society. For a revolution happened when the driving factor became cheap as hell, like when paper was as expensive as sand. It became a real revolution. You need to achieve that. The same thing for compute, for AI compute as well. So you see like 50 or probably a hundred AI chips startups around talking about energy efficiency and whatever. And I know that’s an important thing, but you also have to talk about cost, right? Cost per compute.
Aron Kirschen
But right now, even though it would be energy efficient enough to put an Nvidia graphic card in a smartphone and probably smartphone will be like five times bigger. But assume we could, we would not be able to make a product out of it because nobody would pay for like $5,000 more for something like that or even $200 more. We really have to be like let’s say 20, $30. And I think it will be very exciting and there’s a huge impact on society if we find a way or manage to build a chip. It costs a couple of dollars. Able to run large language models with billions of parameters as continuously power consumption of less than 500 milliwatt. And this is exciting because then we can really talk about live translation, but also about like something like personal assistant.
Aron Kirschen
And if you think about that in the last centuries the only thing people could use to scale up productivity was basically using other humans time. And also of course computer helped a lot with Excel and so on to save some time and increase productivity. But if everybody of us would have a completely personalized assistant, completely understanding that as a kind of default thing in a smartphone, always listening, completely private because running on device then I think we can really enhance human productivity by factor of I don’t know, but not like 20% and come up with like completely new society. And I think that’s what’s needed and I think it can’t be underestimated what happens with the society when we have that. And actually, I think, I believe Zemrun will be exactly the company that enables that. Amazing.
Brett
I love the vision. I really love this conversation and I appreciate you taking this at almost 1am your time in Germany. So thank you so much for taking the time. I appreciate it a lot.
Aron Kirschen
Thank you, Grant. Thank you very much.
Brett
This episode of Category Visionaries is brought to you by Front Lines Media, Silicon Valley’s leading podcast production studio. If you’re a B2B Founder looking for help launching and growing your own podcast, visit frontlines.io Podcast and for the latest episode, search for Category Visioners on your podcast platform of choice. Thanks for listening and we’ll catch you on the next episode.
Twitter
facebook
linkedin