Redefining the Deep Tech VC Relationship with Anzu Partners
|Aired:||March 23, 2023|
Jimmy Kan, partner at Anzu Partners, joins John Cole on this episode of Circuit Talk: Funders and Founders. The two speak about how Anzu is focused on investing not just in emerging tech, but in the tech that will define the next generation. Kan speaks specifically about how much the landscape of deep tech, and specifically semiconductor, investing has changed in the last five years – he sees a bright future, but one that will come with necessary changes to the ecosystem. He and the team at Anzu Partners are working to guide that restructuring.
0:08 | John Cole
Welcome to Circuit Talk: Funders and Founders. I'm John Cole, senior manager on the semiconductor team at MITRE Engenuity. We are a nonprofit dedicated to solving problems for a safer world. Our semiconductor team is hard at work, meeting the nation's challenges around semiconductor breakthrough technologies and the CHIPS Act. Circuit Talk: Funders and Founders MITRE's Circuit Talk podcast and video series, and it elevates the revolutionary disruptive work being done by semiconductor entrepreneurs and investors. This is an exciting time to be working with semiconductor startups.
0:42 | John Cole
The nation is waking up to just how critical they are to our national and economic security. I'm joined today at Circuit Talk Funders and Founders by Jimmy Kan, a partner at Anzu Partners, a VC firm investing in deep tech in industrial technologies. Jimmy comes from a highly technical background. Dr. Kan earned a PhD in material science from UC, San Diego. His first job out of school was working as an engineer at Qualcomm, where he developed new materials for semiconductor devices, metrology and non-volatile memory technologies from embedded AI applications. His name is on 24 Granted Patents and on 33 publications. Today, Jimmy leads a team at Anzu Partners investing in, among other things, semiconductor and materials and computing companies like Afinity, Arduino, nCharge AI, and Liquid Instruments. Dr. Kan, welcome to Funders and Founders.
1:36 | Jimmy Kan
John, thanks so much for having me, and I'm really glad to be here with you.
1:40 | John Cole
Thanks. Well, so I gave a little bit about Anzu. We know you invest in Deep Tech pretty broadly. Can you give us sort of the 30,000 foot view of your firm? You know, where Anzu came from, what does Anzu invest in? How's it different?
1:54 | Jimmy Kan
Yeah, sure, of course. So Anzu Partners has been around for about eight years, mostly as an early stage venture capital firm, investing in this category that we coined called Breakthrough Industrial Technologies. And that can be pretty broad, but it definitely has some overlap with deep tech. But essentially we define it as companies with really strong science and engineering innovations that make major positive impacts in industrial end markets. And sometimes I like to just describe it as technologies you might find in a factory, but for us, the definition of a factory isn't just welding and steel production like you might traditionally imagine.
2:33 | Jimmy Kan
It's also including things like tomorrow's electric vehicle manufacturing, uh, efficient ways to make biologic drugs, and then of course, next generation semiconductors. We're managing a bit north of a billion dollars in assets today across three different venture capital funds, and then also a few later stage vehicles.
2:53 | Jimmy Kan
But I think one of the most interesting things about Anzu, and I think the thing that makes us different is that, that even though we are managing a modest but not a huge amount of capital, we have more than 70 full-time team members split across some investment functions. And then the majority are in support functions. And these full teams that we have allow us to bring not just capital and networks to a specific situation, but it also helps our portfolio companies punch way above their weight class in terms of recruiting top talent, finance, marketing, and communications.
3:31 | Jimmy Kan
We even have an embedded, uh, in-house IP function, and then also an in-house manufacturing and operations function that helps startups transition from development to manufacturing, which has been just massively, massively impactful in the last few years of the supply chain difficulties that most companies have faced. And I think this is really different from venture firms who generally have small teams, a few partners and associates. And because of this, I think we've become the partner of choice for a lot of deep tech companies.
4:05 | John Cole
I was gonna actually dive in just a little bit and say, you know, your portfolio, because it covers some of these folks that are working on, you know, hard technologies or deep technologies, it seems like they need more portfolio support or that that's, that's how Anzu is, is differentiating.
4:21 | Jimmy Kan
Yeah, absolutely. And I think it's about bringing not just the capital, but also bringing the partnerships and the capability. And that's what we've hired our team to do, and I think it's been really appreciated across the, the ecosystem of startups we've invested in.
4:36 | John Cole
So you your, your practice focus is entirely on semiconductors, is that right?
4:41 | Jimmy Kan
No, actually, I, I think in general we are talking about, uh, breakthrough technologies, innovations that make a lasting and positive impact on the world. So essentially it's harnessing technology for its ability to allow us to use our limited resources more efficiently and provide more accessibility for everybody. But for me personally, that means a lot of hardware and then next generation materials.
5:05 | John Cole
Okay. Got it. Well, what are you looking for in startups right now for an investment?
5:11 | Jimmy Kan
Yeah, I, I think, uh, breakthrough technologies are cool or fun on their own. Uh, but as I mentioned, there's some focus around sustainability, there's some focus around efficiency. But, if you're actually asking in terms of in investment evaluation specifically, we do have a little bit of a shorthand framework of some questions that we try to answer positively. And those questions that we try to answer are why now, first of all, and this really speaks to this question of does the technology and the market need fit today?
5:43 | Jimmy Kan
We have this question of do we really, really love this team? And then of course, we also are trying to find out, could we just find somebody to build this over the weekend? And this really speaks to competitive advantage and strength of differentiation and lasting advantage.
6:00 | John Cole
6:01 | Jimmy Kan
There's three other ones that are also very important, uh, but are, uh, sort of following the, the first three. Yeah. And those are questions around is there an obvious quantifiable value proposition? We always have this question of can they even make it? Yeah. Because there are some technical risks that people face. And so we are looking for innovations that need to be really well beyond the science plateau.
6:25 | John Cole
6:26 | Jimmy Kan
And then of course is the serviceable, obtainable market, not tam, but what you can actually capture attractive enough for a 10x return, because ultimately that's what a lot of venture firms are looking for.
6:37 | John Cole
So coming back to some of the technical aspect of things there, you know, you're into your PhD in materials, you work for Qualcomm for a bit, where you covered a lot of ground, the memory and the metrology novel semiconductor devices. You have a background strong enough to do your own diligence on these deals. And that seems kind of rare among VCs, uh, industry and academic background sort of needed to to, to operate on your own there. Why are more VCs coming from the hard tech background these days to do those kinds of deals?
7:06 | Jimmy Kan
Yeah, John, good question because I actually think that deep tech VC work is a lot of technology evaluation and detailed industry landscaping, and that should be well suited to the skills of technical people. But I think it's catching up and now in 2022, the number of financial and corporate investors that I run into with technical backgrounds seems to be a lot more than a few years back. And part of it is probably just me talking to more people now than before, but I think a lot of it is also the industry realizing that an understanding of some of the science and engineering behind these products is really crucial in projecting how long a specific technology is gonna take to get to market.
7:48 | John Cole
7:49 | Jimmy Kan
And as for doing my own diligence, I, I think the technical background helps, but it's not enough because obviously you can't be an expert in everything. And the value chains are very complex when, frankly, you're not just selling an app subscription directly to a consumer. Yeah. But fortunately we have this entire investment team at Anzu who are former physicists, uh, chemis, double e's, aerospace engineers, and biochemists. So I feel like we have pretty good coverage there, broadly.
8:19 | John Cole
Nice. You think that gives you an advantage in deals?
8:22 | Jimmy Kan
I think so. I, I, I would like to think so. But you might be surprised that the advantage isn't so much on technology feasibility or differentiation evaluation or industry landscaping. It's actually, I think much more in the relationship building with founders. And there's something about having just this shared general experience of experiments failing 99 times before succeeding or having engineering designs that were just like, perfect, but the product just didn't work.
8:52 | Jimmy Kan
And it creates some sort of mutual understanding and I think respect about how freaking difficult and amazing it is, what some people are trying to do.
9:02 | John Cole
Yeah. To be able to just connect with them and understand where they're coming from. Uh, do, do investors, like specifically towards the semiconductor space, do investors really need that technical background or do you think it's enough to be able to look at things financially as you mentioned, or be able to judge an entrepreneur maybe interpersonally? How do other investors without that background kinda play?
9:27 | Jimmy Kan
Yeah. I think it depends on stage. So yeah, we do get these companies that eventually get to a stage where it does become about revenue margins, how many units they can ship the supply chain. But if you're looking much more early stage seed stage or Series A, I think it's very difficult to do this kind of work without having the technical background and understanding why a specific product or technology ends up having a lasting advantage over the industry. Because this is an industry that has traditionally found ways to overcome what people have spelled as doom and gloom over and over and over and over again.
10:02 | John Cole
So is it more about being able to sort of filter maybe ungrounded optimism, you know, an entrepreneur comes to you with an idea and they're sure they can pull it off and you want to kick the tires? Or is it about the portfolio support of helping them navigate through it afterwards? Or, um, something else? Like where where does the technical side of this kind of come in?
10:22 | Jimmy Kan
Yeah, I think, of course the relationship part that I mentioned earlier is important. I think the other technical side of this is really understanding the pace of the industry, right? Okay. So understanding that what you develop today may take five to 60 years to get to market - by that time has an incumbent like an Intel or TSMC or Samsung produced something that's better even without this new technology. So there are, it's certainly unbridled optimism about some things that get developed in a vacuum, but it's really about understanding how these various parts all play together at the end of the day.
11:02 | John Cole
Got it. That makes sense. Well, speaking of some of those, um, some of the other rivals that, or potential rivals that you brought up, um, we've seen a shift in focus sort of among some of the, uh, the integrated companies, um, like Intel and Micron and also systems companies like Google and Apple, they're sort of putting, they're putting more of an emphasis on, on power efficiency and on sustainability in general. So not just power efficiency, but also environmentally sustainable process and manufacturing. Can you tell me more about what, how you're thinking about that and tell, maybe tell me more about some of the innovations that are making that happen right now?
11:35 | Jimmy Kan
Yeah. I, I would say the lower power, lower environmental impact, that's an underlying theme across most of our investments. For us, that means we spend a lot of time around clean tech and industrial tech with a focus on, as I mentioned earlier, hardware and materials. And specifically it's electricity storage and generation technologies like batteries and magnets, looking at robotics and automation that can help with recycling, looking at IOT and edge computing and efficient AI.
12:06 | Jimmy Kan
And I, I do think that companies like Google and Apple and the others you mentioned are seeing these kinds of ESG initiatives as a huge priority right now. And sometimes what that means is that they end up becoming some of the largest purchasers of renewable energy and have all of these really lofty and ambitious net zero goals. Uh, I think it's admirable that even in the data centers and in some of the products themselves now, there's an increased focus on reduction and reuse of key materials.
12:39 | Jimmy Kan
For example, one of Apple's recent claims to fame is a hundred percent recycled rare earth elements in the magnets that they use for speakers and sensors, haptics and adapters in some of their devices. And this has really been driven by a lot of innovation in the ability to shred and robotically recover some of these critical materials from old devices. But I think my opinion, and I hope it's shared across Anzu, is that it's not quite enough because the demand for these kinds of electronics and electrification is growing way too rapidly for recycling and reuse to keep up.
13:17 | Jimmy Kan
So we need some really systemic changes like the ones provided by our portfolio companies to, um, to, to get moving forward.
13:26 | John Cole
What are some of the systemic changes your portfolio companies are providing?
13:31 | Jimmy Kan
Yeah. So for example, when I talk about electrification here, we have a company called Niron Magnetics. That company is solving the rare earth crisis by providing the industry with strong, rare earth-free permanent magnets made from abundant materials of iron and nitrogen. Um, we have another company called 6K, which is using a plasma production processed to produce and also recycle and upcycle battery materials with one tenth the environmental impact. So it's not just about making the existing supply chain more efficient, but it's also about taking some of these breakthrough innovations and beginning to utilize some of them, uh, into the supply chain as well.
14:13 | Jimmy Kan
And, and I think that there's, if we talk for a second about the semiconductors and computing side, the sustainability question is also super interesting, but it hasn't gotten nearly as much visibility, uh, coverage, I think as some of the climate change and critical materials recently.
14:29 | John Cole
14:31 | Jimmy Kan
You, you may have seen this, and I, I read a stat somewhere years ago that even in 2015, Google's data centers were already consuming as much energy as all of San Francisco. And then I assume with the explosion of 4K streaming video and social media, crypto mining since then, it's probably gone two times or three times as much, uh, at this point or more. And that's just massively unsustainable from a data generation and power standpoint. And even some of these industry associations like SIA, SRC have actually called out the limitations that will hit in the next 10 to 15 years due to frankly compute becoming a huge fraction of the total world energy production if we don't find something new.
15:22 | John Cole
Yeah. I mean, that, that, that kind of speaks to what kind of impact someone coming out with just a, a lower power chip or having something that, you know, that sits along, sits along those sustainability goals, what kind of impact they can have. Right? I mean, um, you go ahead and knock off, uh, San Francisco's worth of power, electricity, just like that by coming up with a new chip that lowers power consumption. Right?
15:44 | Jimmy Kan
Yeah, a absolutely, and to, to your point, I mean, this something new that we're talking about as well, beyond traditional scaling that we achieve with Moore's Law at this point, I think it's really unlikely that we get another 10 to a 100x raw performance increase in the coming years. But there's hope because there is actually the possibility of changing the denominator of that problem and getting a 100x or a 1000x reduction in power consumption for some applications like AI training and inference.
16:15 | Jimmy Kan
And it, it's hugely exciting for my industry and the industry overall because there's so many great semiconductor companies in the ecosystem now innovating on brand new architectures, uh, ultra low power storage and memory, looking at 3D integration, analog compute, integrated memory and computation. And all of these have a role to play in this coming generation of, of new hardware.
16:42 | John Cole
Fantastic. Sounds like some good investment opportunities there.
16:45 | Jimmy Kan
Yes, absolutely. And we had definitely some of those in our portfolio as well.
16:49 | John Cole
Well, when you look at this, Anzu has to go out and raise money from LPs to back your investments, um, when you look over at your colleagues who are working sort of the SaaS in retail spaces and sort of get envious maybe of how much easier it is for them to raise money from LPs, I mean, how do we encourage more money into these sorts of longer term, technically risky semiconductor, but also other surrounding technology kind of startups?
17:18 | Jimmy Kan
Uh, <laugh>, I, I try not to think too much about it, I guess. Um, but yes, a ton of LP money has been raised and then subsequently blown on crypto dating apps, <laugh> and uh, I guess like get your groceries in 10 minutes kinds of ventures. Right.
17:36 | John Cole
There's a lot of things that there are sort of me too, me too applications of what's already out there, right. Rather than changing the world or doing something truly new.
17:44 | Jimmy Kan
Yeah, and I mean, essentially you're using a lot of venture backed dollars to fund your customer acquisition, which is completely different than what we're talking about here. But fortunately, I think the tide is shifting for our category. Five to six years ago, very few investors, I think were talking about deep tech, clean tech or semiconductors. But it's changed a lot in the last few years. And I think part of the reason for that is people are beginning to appreciate the value of having something that's not only valuable, but can't easily be replicated by a competitor or industry incumbent.
18:21 | Jimmy Kan
And to your point about how to get more money into longer term startups like this, there's no quick fix. It's, uh, it's gonna require an ecosystem of support from investors, from the government, universities, big companies to make semiconductors and other deep tech companies attractive long term. Frankly, the problem is that, I mean, the companies need to get founded and then they need to get enough cash to survive for long enough to get acquired.
18:54 | Jimmy Kan
Then investors need to see that it can be done and then more will survive. And then it's sort of this self-reinforcing cycle at that point.
19:01 | John Cole
19:01 | Jimmy Kan
Cause at the end of the day, it's not really the technology risk as a primary reason that a lot of these companies fail. It's usually just surviving for long enough for the market to fully qualify and accept these changes,
19:16 | John Cole
Really? Do you have an example from your portfolio or something that, you know, just lived long enough for the market to catch on?
19:23 | Jimmy Kan
I'll give an example of something that I worked on while I was at Qualcomm that everybody had been saying was coming next year, coming next year, coming next year. And finally, I think in 2022, we're beginning to see more industry adoption, and that's these next generation non-volatile memories. So resistive RAM, magnetic RAM, Ferroelectric memory, it's taken a very long time, not because of the fundamental technology limitations, but because there's an ecosystem that needs to come together to adopt. So there's various layers of the compute stack where it's not obvious where you integrate this new technology and how it works with storage, class memory versus, uh, cash versus embedded memories.
20:05 | Jimmy Kan
And people need to understand, okay, what are the advantages that we can get by integrating these kinds of new technologies that we're not getting from a combination of everything we already have?
20:14 | John Cole
Yeah. I, I mean, so the other day we were having another conversation with the VC and a, a few kind of hurdles came up and one was simply, you know, there's the time to market, right? But then there's also, it's a somewhat conservative market to adopt anything new. And when you sort of peel that back, it's easy to look at it on the face. But as you said, there are a lot of reasons kind of underneath, like there's a, there's a supply chain, but there's also maybe a whole ecosystem of, say, developers that are sort of writing software for this, and they have to sort of embrace it as well.
20:44 | John Cole
So you not only, you have like a three or four dimensional sort of chicken and egg problem of bringing people over to like a new, maybe a new technology for, for adoption. Right. I'm sure you see that up and down with the other industrial areas too. But semiconductors seems more complex than others in some ways, but
21:02 | Jimmy Kan
Yeah, absolutely. I, I agree.
21:04 | John Cole
At MITRE and the semiconductor team, one of the things we spend a lot of time thinking about and talking about is, is workforce development. So just the number of people that you can have that are qualified to sort of either work at these startups, work at larger companies or, or go back and forth between. Maybe you can talk to me a little bit about what's sort of missing in workforce development right now, or what's missing in the workforce for some of your startups?
21:26 | Jimmy Kan
Yeah, good question. I mean, I think, obviously semiconductor manufacturing and design is something that North America has not had a lot of talent in for, for quite some time with the offshoring of a lot of the manufacturing and process talent. I think, right now, one of the key things to develop is to have young people realize this is a really cool industry to work in. There's a lot of technology innovation happening, and it's beyond software.
22:00 | Jimmy Kan
There are efficiencies that we gain in these 100x to a 1000x types of, uh, architecture developments and fundamental technology shifts that you just cannot achieve by any other means through writing code. And I think that with the CHIPS Act and with some of the other federal support, it's gonna help a little bit, but there's not nearly enough emphasis in those programs today to give us what we're gonna need for the next 10 or 20 years.
22:28 | John Cole
Yeah. And to your point, you know, it is, it's hard to conceptualize earlier we were talking about reducing the power of a San Francisco or two, and it's huge impact. And a lot of, especially a lot of folks coming out of school right now are particularly sensitive about climate change and having impact on that. And just, it takes a few steps. I think we haven't done a good job of packaging that up into a, a real message that delivers, I guess so. Yeah. But, many of your portfolio companies probably have a software component to them. They're out there competing with the Googles and the Facebooks.
23:00 | John Cole
How do you think about, or how do you help coach founders how to compete with them for, say, software talent and other, other sort of limited talent that's out there right now?
23:10 | Jimmy Kan
Yeah, I think this is a very difficult challenge that a lot of technical entrepreneurs are gonna face because you have these gorillas in the room that are paying salaries that are sometimes two to three times as high as you want to be paying as a startup. But I think there's two approaches here. One is sometimes investors will just need to grit their teeth and understand that this is what it takes to hire somebody who's top talent at some of those organizations. Secondly, I think you need to find people that are really passionate about this breakthrough technology or fundamental game approach that you've developed and can have a longer, I guess, roadmap or view of what they wanna be doing in life.
23:55 | Jimmy Kan
And I think it's about helping these people realize that they can make a very big impact in a small organization, whereas in a, uh, 30, 50, 100,000 thousand person organization, they'll, uh, be a lot more diluted.
24:10 | John Cole
Yeah. Maybe the, the, just the concept of joining a small team that can have a greater impact Right off the bat and take on more responsibilities and, and grow their career more quickly. Yeah.
24:20 | Jimmy Kan
Yeah. I, I was actually just gonna give one example of that recently, where we've assembled an amazingly fantastic, talented team of industry experts, and that's at a company that we recently made an investment in called EnCharge AI. And that's a company that is promising to deliver these 50, 100, 1000x improvements in compute efficiency using a very inherently scalable in-memory computing architecture. And this is the idea that you would move a lot of computation from the cloud to the edge and would do battle directly against Nvidia GPUs.
24:58 | Jimmy Kan
And honestly, we were able to get people from the industry, from Nvidia, from Waymo, Google all these places to join this company because of how much passion, first of all the founders had, but also providing 'em with them, with the realization that those approaches that they're working on at these big companies are ultimately not gonna scale long term. And if we still wanna be working in these industries for the next 10 or 20 years, we need approaches like the one that EnCharge is developing.
25:27 John Cole
I'm curious about EnCharge. They're a fabless company. Their differentiation is really in their fabless design. What's their, you know, and you're, you're investing at a somewhat early stage with this company. What is their, you know, if you look down the road, um, are you comfortable just talking about what are the challenges that they're looking at, like right now and they're looking at, you know, in the next two to three years that other semiconductors in that space kind of face, like, what's, what's hard here and what's gonna, what's coming down the road for them?
25:59 | Jimmy Kan
Yeah, that's a good question. I will go back to this topic that I keep on hammering on, which is ecosystem.
26:05 | John Cole
26:07 | Jimmy Kan
I really am firmly of the belief that semiconductor companies are only effective if they fit well and slot into ecosystems where they belong. And I think that we've seen a lot of semiconductor companies and brand new process types of innovations that are really fantastic in a vacuum, but those eventually need to get manufactured at a fab. And somebody like TSMC will refuse to take them on. They can't be packaged, or ultimately, a challenge that I see a lot of semiconductor companies facing is even if they are able to manufacture their chips, have they considered how that product is going to interface with the industry standard tool chains or software that exists.
26:48 | Jimmy Kan
Because really all of these improvements that you say that you're bringing to the industry through your raw technology are somewhat meaningless if the developers out there today can't take advantage of them. So this is something that we are working on at EnCharge very, very heavily to make sure that not only is the technology on the hardware side robust and scalable, but also very easy to adopt from the AI engineer or data scientist side.
27:15 | John Cole
Yeah, that's a, there are a couple of challenges there, right? So like, if you look at going to start to work with, eventually they'll have to work with a large industry player like TSMC or somebody else running a foundry, and they, they have to pick and choose probably where they want to innovate and boil it down to the most, most important innovation and how it adapts around saying around process technology and around all the tooling and, and the fab that they have set up right now. Thinking about the ecosystem, maybe the second problem that they face, they've also gotta go out to the software development community and somehow convince them to switch over everything that they're doing.
27:53 | John Cole
And they, they probably have to invest in a pretty strong what HDK and SDK, something to get them started with examples or something to, to convince them to come over and make it easy for them to come over, right?
28:05 | Jimmy Kan
Yeah, that's right. And I mean, this is why companies like Nvidia and Google have a, a strangle hold on the AI industry, right? Because their tools and their software are ubiquitous, uh, used by everybody at the moment. So I think that there's a couple of approaches. You either take something that's open source that people know about and are familiar with, or you find a way to slot yourself into this, uh, TensorFlow cuda type of work stream that people are, are al already used to.
28:38 | Jimmy Kan
But the idea of developing something that's completely new for yourself, and then you evangelizing that to a bunch of developers, that's a very, very long slog. And I think a place where a lot of hardware companies will, uh, ultimately see their doom.
28:53 | John Cole
Yeah. When you're, uh, talking to fabless companies trying to get, say, openings at, at, at fabs, right? How does that, how do you get a slot on a fab these days? Like, it's, it's tight. We've just kind of gone through this big pinch in terms of chip capacity production. Basically, one of the, one of the key topics that I keep bringing up and we, we want to talk about a lot is just the access to facilities and opening up access to facilities for startups will be sort of revolutionary for giving American startups and American semi companies basically more shots on gold.
29:30 | John Cole
Right? So, and some of that, some of that lack of, uh, fab access as far as we can tell is really, um, some of it is because of the chip shortages and nobody any favors, right? So just, you're gonna get bumped off the line because bigger customers have more established relationships, are gonna come in and say, we just need more right now. Like, you gotta knock these guys out, assunming you do find a spot on the line, but then oftentimes to approach a fab is often, um, it's cumbersome, it's hard, there's a lot of IP negotiation that goes on that goes on. Uh, and so there's, we see, and maybe you see this too, but there, there's sort of a valley of death between when you're, if you can get on an MPW, there are groups that are out there sort of doing that. So if you want to, you know, get a small volume going, you can do that.
30:18 | John Cole
But that next step up is really where you face a, a big hurdle. I dunno if you've seen that this a similar, a similar hurdle or
30:28 | Jimmy Kan
Yeah. I, I think it's certainly a challenge that startups face that the companies like Apple and Qualcomm don't necessarily have. I think there's a variety of different groups out there, as you mentioned, the silicon catalysts of the world that can help startup companies get onto an MPW or some sort of shuttle that makes sense for their size. But ultimately, I think there's a couple of challenges of scale that you're never gonna be able to overcome.
31:03 | Jimmy Kan
You're never gonna beat out Apple for a spot at a three nanometer fabrication run. I mean, it's unlikely that you even have enough cash in the bank as a startup to do something like that. So, what I've found very interesting is that there are a lot of legacy technology nodes. So, I mean, if you look at 180, even 350, 40 nanometers, 28 nanometers is very economical where a lot of these breakthrough innovations actually can be demonstrated and not only show feasibility, but sometimes be better than the digital technologies at 7, 5, 3, 2, whatever you you wanna call it.
31:44 | Jimmy Kan
And so, an example there, actually two examples, are the two companies that are in our portfolio that are working on analog computing. So that's afinity and EnCharge AI. And these are companies where they can demonstrate their technology at 350 or 180, uh, in the case of afinity or at 16 nanometers in the, the, the case of n charge and dramatically outperform the seven nanometer and five nanometer competitors that are in the space because of the architecture and the new technology benefits.
32:18 | Jimmy Kan
So it's a lot easier to find space on a 28 nanometer or 40 nanometer type of shuttle run, then compete with Apple at, uh, you know, 7, 5, 3, 2.
32:30 | John Cole
Yeah. But to that point, you that sort of, it does really raise the bar for the type of innovation you have to sort of deliver if you're a, a startup delivering a new technology, right? So if you, if you wanna reduce the, or go back to an older node, whatever architecture or technology or design you're introducing needs to be, you know, orders of magnitude better than what's already out there, to be able to sort of go back to older technology and still deliver enough to convince a whole community of developers to come over and, you know, all the pieces that we've talked about of like supply chain and production to come over and, and join you, right.
33:07 | John Cole
And you can't do an incremental, you can't do an incremental improvement as a startup. You really have to disrupt.
33:15 | Jimmy Kan
That. That's exactly right. And I think that's where we go back to your earlier question around technology evaluation and why you need to have some sort of technical background in this space to really comprehend what's happening. Because the technologies that do get demonstrated at 65 nanometer or 45 nanometer, or even in these, you know, 180 or, you know, micron size technologies, can they really scale?
33:43 | John Cole
33:43 | Jimmy Kan
So eventually you're going to have to bring that technology that's at 180 down to 90 or 40 or 28. Yeah. Do you see the same types of advantages as you scale? Because even if you're 10 times better right now at a legacy technology node, if you can't scale, then it doesn't matter because the industry will catch up to you otherwise.
34:04 | John Cole
Right. Yeah, that makes sense. I want to ask about a few more things. You recently completed an investment in Arduino. I love Arduino. Um, I started, um, that's, that's how I learned micro controllers. I started out on the app mega 3 28, and then I think that led to back learn the app ML compiler and then, um, was off to the races with a couple different product designs. What did you see in the startup?
34:28 | Jimmy Kan
Yeah. I also love Arduino as a company, and I have a very similar story to you about, uh, learning embedded development on that platform. What we found interesting about the company, and the reason it fits in our thesis is that, well, first of all, the company's super well known and has more than 30 million active developers on the platform. And at the moment, they are well known in education and in this maker hacker community, but they are beginning to take a major step into industrial applications, and that fits really well with our tech, you might find in a factory and edge computing thesis.
35:05 | Jimmy Kan
Secondly, their mission of, I, I think it's essentially making technology accessible and easy to use for everyone that fits super well with one of ours, which is finding unique ways to blend hardware and software together to bring rapid and accessible innovation to the industrial world. So it's not just a cool investment and cool people to work with, but I think it, uh, hits on all fronts for us.
35:33 | John Cole
It's cool among the right crowds, but, uh, yeah, and they've kind of had this, they, they too have had a great story. I think when you look at startups, right? They, they, uh, so their, one of their superpowers, at least from uh, educational development is sort of the open platform and doing an open software platform, so that if you wanted to say, if you roll back 10 or 15 years ago, if you wanted to pick up a competitors, you'd have to understand, learn, pick. And that was hard to get the software to, to begin programming on and developing on so that, you know, hardware developers just had a sort of a free and easy way on with the free software, but they also kind of had started to have this community that came around, started talking, shared designs, uh, and was somewhat welcoming depending on, depending on which forum you went on, they were somewhat welcoming.
36:21 | Jimmy Kan
Yes, yes. The, the forum of choice definitely mattered, but I think it's exactly what you just mentioned around the openness, right? I mean, their entire concept is open source, easy onboarding. There's not this gated feeling of superiority that you can do embedded programming, and the person that's down the street studying design can't. And that was their superpower of essentially tricking people into thinking that they were just doing design, when in fact they were doing embedded programming.
36:51 | Jimmy Kan
And then you have these like different layers of the onion, and you allow people to go deeper and deeper and deeper. And in two weeks or four weeks, people are making really fantastic stuff they wouldn't ever imagined.
37:03 | John Cole
Yeah. Cool platform. So Jimmy, you know, you, you learned, uh, sort of the technical path, the hard way. You went through your PhD, but you didn't take the thesis and spin it up into a startup as we see a lot of folks do. So, but you, but you work day to day with a lot of technical founders. Um, what can we do in the ecosystems to sort of encourage more researchers and postdocs, PhDs to take their R&D projects off the lab bench and out into the commercial world? What's, what what can we do to sort of help more folks do that?
37:35 | Jimmy Kan
Yeah, that's a good question. And frankly, I don't know that I would've had the entrepreneurial chops to do that even if I wanted to. But for the folks that are in school and on lab benches today, I think it's about access to education. So these guys have tons of education around, uh, AI compilers and micro electronics fabrication and crystallography, but frankly, there's not much exposure to what it means to raise venture capital. What are other sources of funding?
38:07 | Jimmy Kan
What types of roles and careers are out there for them besides just picking up a job at a Qualcomm, Intel, Micron, or, you know, you name it, other big company.
38:18 | John Cole
38:19 | Jimmy Kan
We should definitely have the increased accessibility to the accelerators and incubators, but I almost feel that we need to go one level earlier to bring this, uh, concept of entrepreneurship into people's engineering education as well.
38:34 | John Cole
Yeah. Well, you, you know, the, um, something we see time and time again with technical founders, uh, it's oftentimes they, they fall in love with the technology, not the problem. And, and sometimes like looking at things from the customer perspective is can, can be kind of a challenge, right? Rather than just sort of look over the technology at where it's being delivered. So agree. But if you sort of look up the educational, uh, supply chain or look earlier on to sort of teach those lessons, um, could, could have folks coming out with more entrepreneurial mindset. Yeah.
39:05 | Jimmy Kan
39:07 - 39:13 | John
Jimmy, what's your message for innovators that are trying to bring something new, uh, to the market and semiconductors right now?
39:15 | Jimmy Kan
Well, first of all, I think this is one of the best times ever in the last few decades for hard tech and semiconductor innovators to begin entrepreneurial endeavors. So don't be afraid. There's a lot of capital that's out there that's finally willing to fund these sorts of advancements, especially as they relate to, um, sustainability and improved performance. The other thing that I would caution and advise people to think carefully about is I, I've used the word ecosystem many times today, and I'll emphasize it again here.
39:55 | Jimmy Kan
For a lot of deep tech companies, and especially in semiconductors, your innovation will only see the light of day if it brings some sort of disruptive advantage without too much disruption to the customer. So think about that ecosystem, think about that supply chain, how the customer and ultimately the software player at the end of the day, experiences your, your technology.
40:20 | John Cole
Great. Yeah, that's really, that's really helpful. Well, Jimmy, thank you so much for coming on the show and talking to us about semiconductor technologies, talking about your investments and talking about the ecosystem. It's been a great conversation and can't wait to have you back.
40:32 | Jimmy Kan
Yeah, John, it's my pleasure. I can't wait to be back. And thanks for your, uh, your time today.