In-Q-Tel on Chips, CFIUS, and The Valley of Death

Transcript

if we all agree that this is an important technology right important enough to block and then it should be important enough to protect and important enough to support and I think that is the thing we're really missing right now Yen Jung invests in microelectronics for inel the cia's Venture arm yes that's right the CIA has a venture arm named after James Bond's Q I don't like it's both cool and cringey I'm going to go for it Yen has a PHD in electrical engineering spent time at DARPA and on the hill as a science fellow again welcome to chinat talk hi Jordan thanks for having me so what is inel sure yeah oh inel we so the way you described it I think we we would May re redefine what you said we are associated with the CIA we were founded by the CIA 21 years ago and I think a lot of folks think of us as the Venture arm for for the CIA but in reality we are a nonprofit uh organization and as such we actually support more than just CIA at this point we actually support eight different agencies and within not only the int Community but broader in other organizations that deal with National Security and so for us that really means is that we think of ourselves more of a strategic investor and I think there's an important distinction especially in the field when you think about what a large strategic investor would invest in versus what a a venture capitalist would invest in and and some of their motivations and specifically for us we're more interested in helping our government customers find the right Technologies and be able to leverage those Technologies at times that might mean that from a business point of view this is not companies that we select might not be the best business opportunity but we look at it from the technology and from the mission and use those as a broader scope and lens in terms of deciding what we uh engage with so how do you guys split up your time between talking to your clients within the US government and running around San Diego trying to find the next hot microelectronic startup yeah we so we were pre-co times we would have access to go into into the government facilities to talk to our customers being the fact that our headquarters is Virginia so that's where I am based a lot of our customers and and the government are in that in that space and having that face Toof face talk is is important to better understand what they're working on we also have two other offices in the United States one in the Boston walam area to really engage with some of those innovators and entrepreneurs coming out of you know the MIT and and Harvard and just that the scene up there but also we have another office in mland park where that's where most of our investment um Partners live and and obviously that gives us access to a lot of the Venture Network so I think geographically that kind of helps us situate not only with entrepreneurs in the west coast but also with a lot of our government folks here on East Coast to help drive and understand better what are they interested what what what problems would they like to solve and that that you know evolves every year right as technology and use cases and situations change their needs change as well as the situation awareness we provide them changes in the sense that new companies come up all the time and new technologies and we want to make sure that we are relaying that information back as well say hey this might not be on your radar but maybe it should and let let me tell you why so we had Mike Brown of diu on the podcast in late December there are a number of figuring out how to bring private sector technology into the government is a long-standing challenge I'm curious how you guys see yourselves in relation to the other government Affiliated folks who are also running around Silicon Valley yeah no it's it's great this is such a an important area to be in we think that the fact that others are doing are looking at it from different perspectives I think is important we obviously feel like we offer a a specific perspective and capabilities that we think the government should be you know leveraging even further to to be able to get the Technologies and that they need for their missions what mik brown da is doing is great and the way they interact with with startups look and having them help look at the dod side of the the house of course we we look more broadly so we do have parts of the dod as as our customer base but you know also Parts on in the intelligence and also just law enforcement type of um group so we we have a little larger aperture right and then our the way we look at Technologies is slightly different there's also the way that we can actually make Equity based Investments I think is it's one of those key differentiators that that we have in a sense that that kind of gives us the ability to act like a venture capitalist or strategic event investor sit in the same kind of room board rooms and have that uh discussion with those Folks at that level which is which is where the startups are playing which is where kind of their boards are telling them what to do and that situation awareness is I think important and but I would be remiss not to mention other folks in government like what aworks is doing in the Air Force we we really like those guys we think that the way that they are leveraging the sbir program SB is the small business research Innovation program it is a program that for organizations or agencies that have a research budget um a percentage of that of that research budget is taken off the top and work towards these grants for small specifically for small businesses and I think the government at this point defines them as 500 people or less there's also certain categories for I think it's like a veteran owned female owned in some of these special categories that there's preferential treatment for evaluations but organizations like DARPA NSF NIH right lots of those organizations that have a research budget have a separate office that then takes that that cut off the top to to deploy onto um small businesses there's been a lot of talk I think in the past of are these programs interfacing with the right folks are they interfacing with the right startup entrepreneurs that with entrepreneurs even interested in doing sbir right if I Venture back funding would I even want to mess around with doing government grants and and auditing and all the stuff that's required for that or is that program made easy enough for me to do that with low overhead so there's a lot of kind of operational efficiencies I think that could be that could be had and I think a Forks is doing really well in that and also giving them opportunity to find government customers and talk to them there's a we could go into that but I think they're better equipped to discuss that but I think coming from uh as you mentioned when I was at in obviously has a SBI office as well and we fund a lot of um really Innovative kind of startups but it's also ones that been around for a while and I think the pajora of word be spr Mills right for folks that just don't have that intention of actually turning into a growth company but more of just looking taking that money to to fund something to do it again and go through the whole um process so we want to be supportive of the ones that are want to be growth companies want to take something to into the commercial sector and those there's there are ways that such as incel identifying those companies and then helping them grow I'm from the government and I'm here to help in Silicon Valley has been a tension point to say the least over the past I don't know say decade or two what is your sense on the sort of relative willingness and resistance within the startup ecosystem more generally to get involved with with government projects and I'm curious if there's any difference on the software versus Hardware side yeah it probably speak more on the hardware side than the software side um predominately that's where I invest in I would say that this there are concerns of government but perhaps not in the ways that most people think I think most people think about like Google's Project Maven kind of situation I think there there are a lot of startups that do recognize the government as being an early adopter right an early validator of certain technology especially in the really hard tech areas you really do need someone who understands the regulatory landscape or who has the ability to actually make a large capital investment so it doesn't even have to be defense related you think about energy energy related programs what maybe what Bill Gates is trying to do with restarting nuclear you really do need to have government involved in that right there's probably no way of doing that investment without having those few Folks at the table and and so I think for certain Fields orally in Hardware there's a recognition of that right you have for instance a lot of the AI chips companies out there do you recognize Facebook and and Amazon might be partner for their large data centers but by the way government also has large data centers they could be customers so let's make sure that they're involved this well so that's one I I think there's a there's another concern that they do have and when when we say hey we well we not from the government for one if we could get even B the ear of the government I think some companies would always say hey tell us about what's going on with this regul regulatory thing what's going on with cifas which hopefully we could talk about later in the segment what's going on with with China I think now there's a recognition from a lot of the folks in the startup community that at least have to be up on what's happening in in policy world because that might trickle down and affect the way they do business having a hand off and and approach might not be good for business strategy going forward yeah it's interesting sort of thinking about this in the context of the Chinese system in military civil Fusion where presumably whatever the inq tell is doesn't get like people don't turn them down a lot I I it's an interesting feedback loop right where the government the the sort of like defense and intelligence communities have to be good customers to win over the support of to just catch the eye of these folks in the first place whereas in a different system where the sort of like Tech transfer and you take this contract and you don't really have a choice is more of the dynamic than it's well anyways you Riff on that y yeah I I think you're right we pride ourselves on how hard we we work and we actually have a full set of folks that whose job is to be that conduit into the government to say is this working out for you it might are we going to successfully transfer this technology that we've invested in and make make sure that that actually gets used in the right way give us feedback CU a lot of times the customer not only a customer the company wants the feedback right this is when you're starting a company that feedback is probably like the most a lot of folks are willing to actually sell a product at a loss if they could get feedback first that the initial kind of is this working for you are we making the right mests because that drives this second generation and if we could catch them at the right time we provide some good Insight from our government customers like hey this worked or that broke because I put it in in in this environment and then and that's good insight for for for the company and for customers as well and really to to you need to have folks to really engage in that so we actually do it's Hands-On and I think a lot of folks that when I was on the hill the folks that the the old-timers I think that were working on manufacturing policy for the last 20 30 years would say that this type of work is a contact sport right you need to be able to be there on the shot floor talking to the folks and saying hey how can we help you what's going on here so we think of it as as the same thing we we are very Hands-On and on the solution transfer process there are ways that I think if you look at what you China is doing that I think they example that comes to mind really is hsmc that's really that recently came up and for folks that aren't familiar this is the they should be if they subscribe to the chinat talk newsletter at chinat talk. substack do.com but anyways continue in there you go yeah so this is a a just a semiconductor debacle I think and it really exemplifies what you were just saying Jordan that that with China's push towards getting all Hands-On deck government and and civil and everything within China to become semiconductor independent as and throwing money and and bankrupt ing any effort to actually get to that goal has led to sometimes a a unmanaged expectation of how to actually get there and Hing semic conductor hsmc in Wuhan is example where you know a a charismatic businessman over promised and didn't quite understand the complexity of building a state-of-the-art some connector facility you know they quite understand the guy didn't have a high school education yet it was the con of a lifetime he he built a billion dollars out of out of the local out of the local Wuhan government that's right and not only that he was able to what I thought was very striking is that he was able to recruit one of the top R&D directors at tsmc and so this is and and you could also argue well how is if if there's discussion of like how protected is tsmc technology and IP and folks from China but the fact that this non company right or almost the fake company is able to recruit a a R&D director from that organization as it's one concerning and and two the the fact that they were able to get a tool one of the key ways that us has been able to slow down Chinese development in some of these key areas is by saying hey you're going to have limited access to tools one of the tools there would get not a Us tool the Dutch tool it's called as by a company called asml makes the premier theography tools the these are the tools you need to actually print out the the semiconductor designs they were able to get it right and they use that as collateral to get more money to fund the facility but the fact that they're able to get their hands on this machine was actually concerning as well let's talk a little more about the Chinese sort of Chinese chip Ambitions smick also has a ton of X tsmc Engineers this is not like plenty of like more legitimate companies have been able to to recruit some of the best talent in in in in Taiwan onto the mainland it's not just the not just the scamers who've been able to do it to to what extent is a rising Chinese chip industry something that the US should be scared of in the first place the current discussion now is on supply chain security and making sure that if there's an earthquake or a war in Taiwan the world doesn't come to a ing halt but what also to what extent is like fears of becoming a more and more relevant Global player something that the US really should be all that worried about yeah and I think you're you're seeing this like in just in general that the Investments that we're talking about from the policy side of onshoring a lot of the semiconductor supply chain to the US is partially response right to to Innovations you're seeing abroad not only in China but tsmc and Samsung how much how much of the market share that they they hold and how many of us chips have to go through those kind of critical points you know I think it is it is one of those things where as a technology person it's great that Innovation is happening across the world we do want Global Innovation because I think in general everyone doing well is good right from technology development exploration Discovery perspective on the other hand I think looking at what you China wants to do in terms of becoming a a self- sustaining self indpendent entity I think that's we do need to try to understand you know how that how that strategy might impact where the US is what it's important to look at that this is the way that they're doing things aren't isn't anything new right we've been in the situation with when with Japan right in the early 80s and their rise in semiconductors and then later in Korea both of them actually took a very similar roadmap to get to where they are and mean that they really started with memory and so Japan the Japanese kind of uh semiconduct industry really looked at hey what we want to get into this thing called microelectronics semiconductors what what should we start with first and you memory ended up being one of the things that they started out doing first by Intel did that as well because it's one of those kind of easier things to do memory cells you stamp them out over and over and over again so the process is not as difficult but it gives one the ability to learn about okay this is what I need to do this is the things that I need to do to like increase yield and and in performance and things like that and you saw the Japanese uh industry use that as a stepping stone to get better and wiser at this process and now you know that to get you where they are the Koreans have done the kind of s similar thing used memory as a stepping stone if you look at Chinese kind of facility Investments a lot like we look at the hsmcs and maybe the SMI the UMC type you know Investments but but really A lot of them is actually in Legacy old you know facilities to do memory a lot of it is just like memory stuff and and so you really have to think that yes they have this shiny object that they're really trying to tout and say hey we're really top but at the same time there is a big push also just to make sure that we they understand the technology and they could get the basic which is the memory part down there are a lot of ways I think we should be thinking from The Innovation point of view on how to what should we do in in terms of that strategy and and we could share about things that we we think that the US can be investing yeah before before we turn to your policy uh proposals it's interesting to think that given the way sorry it's interesting to think that like China yes like they're working on memory but they also have been able to create worldclass design firms like high silicon and and and Hami and whatnot and that was not really a path that was necessarily available in the ' 80s or '90s which you can like start working at a very different part of the supply chain at the same time as you're building out your manufacturing processes yeah that's right and let's looking at a recent plot of private Equity investments in China and everyone even thinks about the supply chain side of them making physical stuff but I believe it was more about 60% of the private Equity is actually going to Fabulous companies in China and and 20% is going to actually manufacturing so they are actually pushing into making sure design because that's where the value that's higher up the value chain right if you look at where the US is the top fabulous companies are us companies right think of the Qualcomm the broad they don't they use other facilities to to manufacture physical chips but the IP is what where they make the money on and that's that's been a model that's worked and I work I say worked in with an asterisk because although that success of that model in the last maybe you 10 10 15 years has created some of these large companies you seeing the fruits of that right now right where an over Reliance on this kind of fabulous model has also really allowed us to relinquish a lot of our manufacturing capabilities and if we you know one can argue if we kept what we were doing and and rejected support early on we wouldn't be in the situation we are now where we have to probably spend a lot more money and then training and Workforce and all that stuff to bring they bring that Innovation back so so there is something to be said about having a balanced healthy uh ecosystem rather than a more lopsided one even though from the efficiency point of view that might be it might make sense yeah we're recording this on April 9th 3 days before the leading Lights of the US semiconductor ecosystem depend on the White House to talk about the ship chip short and they need to fund the chips act and whatnot the biggest ticket item that came out of that that legislation was money which is likely going to go to support Fab Construction in the US and you and some of your colleagues at inel put on a really interesting paper that pointed to some other parts of the Innovation pipeline which the current way that Congress is thinking about spending money doesn't directly address so maybe before we talk about the the recommendations in particular why don't you walk through the sort of like stages that go from research all the way to to PE investment and and IPO in the in the chip space and where you what you think is missing in the American context sure yeah I think we think of this commercialization pipeline is really starting out with key Innovation key research and that has to do a lot a lot of us funding for you know basic R&D right and and that we should say has been stagnant and or by and and some I say has dropped as in relation to where China is doing in investing I think China might be even surpassing us within this decade in terms of R&D kind of R&D um dollars for research I think that really starts off a lot of this lot of innovation and then you have kind of organizations like the daras and nsfs and doe office of science that really fund a lot of at least from the microelectronics side some of that some of the research that for what companies call pre-competitive research right this is basic research on how do you build a semiconductor what are the new materials we should be looking at those are all kind of important things as we look at the future then you get to when you have you past if you're used to the government jargon you have pass the the baa stage right these Broad and agency announcements which you know these for Grants and such you can get to the point where okay maybe I as a small company I'm I'm using research from let's say an NSF or DARPA kind of grant that I got me a few years ago I'm ready to turn turn this into a company and this and then that's like a weird kind of transition point and it's actually a point that we think is is there's actually quite a big gap and in fact if I you just take a pause here and mention that incel is as as an organization is not just we don't just do invest Investments we actually have much several different divisions that that actually look at part of this supply chain or this commercialization pipeline one of them is a new group that we just started called emerge which actually looks at how do we help R fundamental R&D get out of the lab and into the commercial sector and so let's put putting that in context so let's say I a researchers at a a a federal research lab right when our discussions with some of those commercialization directors a lot of them say hey the biggest Gap right now for us is actually not not as much resources in terms of funding although they probably after now you said that they'll probably say no we want more money but a lot of is actually just human capital right they actually need the right people to help guide these companies the the greatest number they said of companies that didn't make it are the ones that didn't even start and what I mean is a lot of these you know Innovation research resarch hey I just maybe patented this really cool thing but man looking at how am I going to take that out of loud you take a pause of my career I need to do all these things I'm not sure I'm ready to take the leap or I don't think I have the resources to allow me to do that and therefore I'm just going to keep that on the shelf right I'm not going to to mess around with it and so that if you if you take that context we're actually losing a lot of innovation because we perhaps are not supporting these researchers on how do you do business formation what are the legal implications how do I find a mentor that knows that say semiconductors well enough right there's probably a lot of there are a lot of mentors and accelerators that do software I was about to say let's stay on that for a second like I am sure this is not a problem in the computer science PhD ecosystem there are hundreds of DCs running around those labs trying to get these folks to start startups so what is it about the hardware space in particular that means that you have to have inel not a not a fully private company fill fill this Gap yeah yeah and we we actually spoke to a lot of these incubators Tech Stars and you know Y combinator and and you're right A lot of them are they're known for SAS products and software software products and a lot of their entrepreneurs and residents and mentors come from that ilk and then so when they see a physics-based kind of company that's hey we're doing this Nano Crystal thing they feel like I could maybe help you on talking to some lawyer or like a guy to get a loan but in terms of how you should formulate your company and which direction take the technology you usually need a subject matter expert that you just don't have as many of those people and I think it's not overnight thing it's been a a longstanding shift right when you and from The Venture community of moving away from hardware and when you move away from Hardware you get people who just are are leaving that field right the expertise is doing link to the workforce that that you need to to do that uh you just don't have those as many of those people they exist they just not as many of them as you have in software and it's very hard to to start a company and I should make a plug for companies like Silicon Catalyst in the Bay Area those are seasoned veterans in the semiconductor industry they've been around for a while and have been and are helping companies in the semiconductor space with access to certain tools and and networking and things like that there's also Elon G's organization ation select cyron road that helps organization helps the folks in the research Labs Government research Labs meet and live in the Bay Area so that they could actually start to figure out how to turn their hard tech kind of innovation into something so there are groups that do that right uh but there are few in far between and much less you know well funded than the really big kind of organization that most folks are you know familiar with so um in the this paper I came out recent I I wrote with Danny and Chris recently on labs for Fabs we had a whole section on labor development and as a an an electrical engineering PhD I'm sure at some point in your career crossed your mind that like life might have been a lot easier and uh more lucrative had you gone down the computer science path do you have any thoughts more broadly on education policy or the way the aside from this stuff just like more like broader things on getting folks more interested in pursuing careers in this the it's pretty dramatic when you look at the numbers of how much what what percentage like immigrant first generation immigrants fill up the the the graduate programs in the US when it comes to electrical engineering yeah yeah that is very true and when I was in in grad school I actually started a a whole course series on like engineering an engineering course on how to get folks who are non-engineers to think like an engineer so like the engineering design cycle and part of it was was kind of saying hey there's like you said there's lot not a lot of folks that are getting into this field and getting their hands dirty making stuff anymore what can we doing it the way that we did it was more of a community based approach and by no means am I the one that innovated it I I took a a lot from from the ties program ecsd and they took this program it was the Duke that a lot of in Purdue and a lot of those organizations and univers been thinking about how do we in how do we make a more human face to engineering and that I think in the past folks be like you can make a bunch of money if you were if you are an engineer or software developer something like that but if you look talk to the the folks kids today that are in this field that's not the main driving motivator right a lot of folks want to make a difference especially in the context and in the environment we are today right there's a lot of Need for for Change and actually impact and that's something that they're looking for more I think we could be you know taking taking a queue from them and saying look there is a way to use engineering as a way to impact community so what we did was work with local nonprofits so this this is a local zoo in Santa Barbara we say hey what is your problem can I get engineering students to help you solve a technical problem and let's let's let's get some uh you know funding from local community to do that and actually they have a little kitty ride that's like a zo a train and sometimes the train actually derails which is right from the heat um so we had Engineers hey let's how do we actually put a monitor on that and monitor that and have like telemetrics to say to the mainten staff hey this rail is actually starting to increase because of the weather the heat let's preventively maintenance that's something that I think is kind of neat cuz you're helping a community and you're practicing your skills and along the way you're going to learn about hey what chip should I use to do this like how do I should do this telemetrics and maybe I could be using this information down the line in my career so you I think having the human face involving the community is is one way one way to do it and I think we need to stop saying how much money we'll be making because I I don't think that is the right motivator for anyone who wants to get into this field yeah I I was not expecting us to to end up on zoo trolley trains when we started this conversation but it's a really interesting angle because competing on you just given the way that software scales and the sort of nature of the industry like it's going to be hard to match the salaries out of college that you can make as an electrical engineer versus one you can make as a computer scientist I just read I'm doing an interview this afternoon with uh a guy who wrote the story like the early wrote up the early story of of SpaceX and book like the level of like the level like how much cooler is it to build rocket ships than than a recommendation algorithm and I think that's something that when the sort of advanced manufacturing Universe in the US disappears like Miss that of building real things is actually incredible and magical in a way that algorithms are cool also but there's something different and tactile and with material like Material Science which is something we're going to talk about in a little bit is just so unbelievably fundamental in a way that the Innovations of the past 20 years of the internet of the consumer of consumer internet just aren't and there's something really exciting and special about that and I think Elon Musk for all of his faults is a incredibly powerful example of how big an impact on the world you can have doing real hardware and hopefully I'm sure he hasn't had the movie hasn't been made about him yet but like whenever it will be it'll be much more exciting than the social network to be sure yeah I think I think that's to me that's a a great reason why we should make sure that we're supporting fundamental research because you know who's funding some of the early stuff that SpaceX was doing NASA and and government organizations so much like how the US micro electronics Industry started from defense department funding and and so again I think we we can't lose sight of the beginning of this pipeline is fundamental research and then those research programs out of n NSF DARPA NS you know NASA doe are the ones that touch the students first not us the they're the ones that go down to the undergraduate and graduate level that get folks excited and then those these are programs that have high school students come in and see the research that you know uh these guys are you know guys and Gs are doing to get FKS excited about pursuing a career in science and and Engineering yeah I completely agree and I think there's a large government component in in terms of the research Community that's involved in that coming off of that Yen the sort of nature of law has led firms to think in the an 18 to 24mth Horizon because within a relatively short time frame you have Innovations in science which like totally break the Paradigm and whatever you were making is Obsolete and like way less valuable than it was before so given that nowadays we're seeing mors law slow down and those Horizons and those timelines between nodes increasingly expand and become you know way more expensive what role should the US government have in trying to rejigger the the timelines and Horizons of Industry to better to better line up with the new technological Paradigm that we're increasingly going to be U stuck in yeah oh that's a great question and I could answer probably in two different ways I maybe stting the micro and then com go to the macro and hopefully link it to the infrastructure kind of the debate we're having right now and I'm going to speak just maybe just on my little wheel house which is in micro Electronics obviously there's you know a lot more we could be talking about in terms of this topic it it's first we should recognize that making some facil in in fabrication that is is a very wasteful process right it's a subtractive process meaning you you're really just removing semiconductors using nasty chemicals and things like that to get the parts that you want and there is partially a reason why a lot of those facilities actually left in the United States let's not forget that the what we call doe super fun sites right these are super contaminated sites one of the largest density of super fund sites is in Northern California where the early semiconductor facilities are and so that is there's a reason why I think with lack environmental regulations and things in other countries that facilities thought okay why don't we just move those over there we pay less we know to worry about cleanup and so we have to reckon with that we were exporting not only some of the knoow but also a lot of the the the waste as well so I think that's to to True your point like it is we do want to make sure that we are more if we're getting a second shot at this that we're more mindful of the environmental regulations environmental kind of harm that could be done and let's plan for that ahead of time to be more sustainable and so using sustainability in environmental context here but at at the same time from a technological point of view instead of making chips that you throw throw the whole thing away there is this idea of now of um modularity it's coming back right we've done modularity many different times in the past now that we we've been on this binge of saying like the whole chip itself is going to be Strunk better and better and we focus all of our attention on the chip itself as we're looking at the the end of that road a lot of folks are looking at what was called Advanced packaging as a way to prot potentially and not to reuse re redo everything but we could reuse components what I mean by that is what if I just make a chip for 5G and make just that 5G portion and then I put on the shelf and say I'm going to reuse that part later and then I have a memory chip and said okay I'm I'm going to have a big set of memory chips and I'm going to use that and a different time and put it together like Lego blocks instead of having to say all right I this one design I'm going to combine everything together and bake it and it's going to be hard baked and that's it and then you find out a year from now oh a standard changed I just throw the whole chip away and start all over again with a more modular approach you say okay the standard changed I can't use this memory block anymore that's okay I have another one on the Shelf I'm going to plug it in and hopefully I can reuse the other parts of the chip the the fabric the base and everything like that that that's a con that's a thought at least because of that more modularity and the challenge now is how do we make that modularity perform better than if you were just to do the morw approach because I think that's still a key driver a key metric is performance so there are many different ways in the advanced manufacturing kind of push people are doing to do that in terms of interposers and embedded Bridges you make sure that you're doing your highspeed Communications and so forth but so that it is one way that I'm looking at it on the the larger front I think when you're looking at infrastructure I think we have a good opportunity to look at how expensive it is to actually rebuild new and it might be better to just maintain along the way and I think there's a really good opportunity for embedding smart sensors to help with that process can we detect cracks and there's there ways of useing fiber optic to look at stress fractures and things like that along really long runs of tunnels and bridges can we make make a little smarter kind of investment in some of the smaller sensors so that our larger infrastructure is well maintained that's another way of looking at sustainability and maybe how microelectronics can be part of that in this larger context coming back to the discussion of where the Innovation pipeline is stuck in the US one of the proposals that incel is pushing forward is the need for a US investment fund why is that why why do you think of all the billions of dollars that us could could be spending on chips why do you think this makes this makes more sense especially relative to the the checks that are probably going to go out to Intel in global foundaries in the you know foreseeable future yeah yeah in this case we're really thinking about a fund that is is helping the little guys right the smaller startups not the large companies I think they have they're okay by themselves so going back to this this idea that in in this Innovation pipeline there's a lot of kind of inputs that one needs the resources and and one of them we discussed was R&D funding in we established that look that has been stagnant and then one of the that being one of the key funding resources that a lot of companies have used in the past to get to a point of commercial you know pre-commercialization is are you know those are those kind of Grants another area of actual of investment poost that is the Venture funding right by no means is that the only one there are loans and other things that one can do what we're talking about today is me more about the Venture community and US Venture firms in the hardware section especially in semiconductor tools and maybe the non- AI chips like more currently unsexy kind of chip designs they are not receiving the the investments from the community right as you've mentioned they've there there's a big push towards movement towards software and then the the ones that are left in Hardware really picked the prized Bunches of the litter and and then you the ones that are left over they're so good Technologies but you just don't see the funding necessarily for growth and then the Third Kind of things that we're looking at is is that the US has actually restricted a lot of foreign investments so in the past a lot of foreign or entities have invested in US Innovation because a lot of innovation does uh start and originate in the United States and companies in China like China have invested in those but with with there's recent restrictions such as cifas that has restricted that that's another kind of source of Revenue or of that of capital that companies are are not receiving we could argue whether or not that's a good or a bad thing but I think it's more it's more important to really just recognize that as a reality that is one of the thing that they're contending with so then the second question is that what do we do about so yeah so well let's talk about sort of cifas and the second order effects and then come back to the what an investment fund could do to try to amarate them with these three kind of areas of lack of funding from different resources where we we think is that look like there's there's a need for these companies to to be successful there's a lot of we've already identified these technologies that have just not gotten Government funding when yet the Technologies is important to National Security right there there's important to actually techn technological development and there are kind of discussions with a lot of larger large tech companies they said that look we actually invest in funds outside the United States these kind of these more Sovereign you think of them Sovereign wealth funds I think like in Taiwan there itri right they're taiwan's Institute to to support the semic conductor industry there they have a fund that they use to support some of the startups and businesses there Korea has one for for their companies and in Belgium if IMC which is a semic fabrication facility that's open to to Global use they have a their own fund to help support those technology there are examples and there's other types of examples but the point here is that the companies we've talked with the us-based large corporations with Strategic investment arms they say look we deploy our dollars globally across these things because we could we see Innovation potentially coming out from Taiwan or Korea or Belgium or wherever but that the opportunities to use those same dollars in the US is limited if not there's none and our argument is why don't we have something large enough where we could deploy against some of those key National Security kind of interests in better tools for semiconductors for preparing ourselves for the new wave of of postm law semiconductor equipment that's going to be needed do we what kind of how we should be set up we should be setting ourselves up so that we have the capital to be able to do that and and the people ready to do that we just don't see that there and so we say that's one of our recommendations that we think the US needs something like that I think this is going to to cfus right this commit on foreign investments in the US and so this is multi- government agency committee that literally looks at mergers Acquisitions things of that at a certain dollar amount right so these are typically it was at the larger dollar amounts of and making sure that us companies aren't getting you know undo influence because of of certain merger acquisitions by foreign entity and specifically certain entities like you know Russia and China things like that from a lot of the startups point of view like that the process is a little opaque they don't know when if you're going to be part of you know part of that cfus review and if you're in a technology kind of space that might be encroaching on something that they're interested in literally everything yeah it's it's it's partially because people you don't know they don't have an idea of what that might be and so you go think of that as a kind of like a tax that that then investors say there's risk now there's a l higher risk so then we're going to take a higher value a higher you know Equity stake or something like that so from the co-found from the founders point of view that it's not great but at the same time look it we do need some type of those protections it's just that one of the the unintended consequences is that we are the policies right now is all stick right you can't do this you can't you can't fund you can't you know get this money or you can't do that and what what happens that with all these barriers that are just constructed for these startups they're looking around for funding and a lot of some of the US inv investors say look we don't want to touch this sorry we can't invest at you and then what's left is a Chinese says look why don't you just start up shut everything down and and start a shell company and then give me your IP we'll give you we'll give you a much lower buy it on the cheap but at least we get some money off of it you get your investment back and and reluctantly we've seen companies do that right like they love this technology they think it needs to survive and do something and there's always a hope that maybe just give it a little more dollars and give a more love it will come back in in in a great way but really what happens that the Chinese are and mostly the Chinese as as they have the money they're able to get this on the cheap and let's not forget a lot of these Technologies had because of this pipeline we're talking about have a long history of maybe us R&D funding right so this could be tunes of millions of dollars that they've gotten to get this technology to the point where they are now and so not only are we losing technology but we subsidize R&D for a foreign entity and so you know that that kind of the unintended consequences and what we're saying is that like back to the investment idea is that why don't we have a carrot along with the stick right so if these are if we all agree that this is an important technology right important enough to block and then it should be important enough to protect and important enough to support and I think that is the thing we're really missing right now is that okay we said that this is so important that we're going to block it how are we going to make sure that we we give it the resources to to survive and to be fair to the commit doesn't have that power right we there's no entity right now that can say let's inject some funds right this is a a very kind of difficult situation probably requires a lot of phone calls from different entities to figure out which vehicles can we use and is if folks who know government it's a tall order to really ask for and and this is not just we don't do this you know every day this is there could be hundreds of companies right like in this especially as the bar for cfus has actually gone lower that they've got the expansion of that strengthening cus has actually opened up the aperture more that I think we're come to the point where okay where are the carrots and how do we actually make sure that there's a a a a suitable and sustainable mechanism to give these carrots not just say that we they should be supported but that there potentially is a fund that they tap into from after review that hey okay we said you can't take this funding here is some to get get to get started and by the way that could help ease a lot of the risk from the US Venture side say okay if they're going to get funding from this organization I see that there's going to be government support for it it's not going to go away the we've the government in a sense have bought down risk and that would help rorate and at least get some more of the US Venture folks back into the game and so I think what we really needed to show is that we're going to take the leap and we understand the pain points so let's eleviate those another one of the ideas which you guys proposed was this idea of a micro Electronics sandbox why is this so important particularly if you're trying to support new crazy nons silicon or posts silicon materials research I think that the point here is this is the one of the key differentiators what we talked about from software to Hardware Hardware you have to make something right you can't you can go hat in hand and say hey this is a a a sketch I put on a napkin of a chip now I need $10 million to make this chip and and so for for many Hardware Investments you really need to show an early prototype some type of we call tape outs where when basically you send the design out to the Fab you need to get to some Milestone usually it's a tape out or engineering samples that can be expensive and if you're especially if your startup that you're going to be designing at 16 nmet or even 7 nmet that could be millions of dollars that that you need to do that and so there a need to have a ability to kind of prototype to actually make some of these things to help buy down risk that's one that two is that all too often a lot of in a lot of researchers again that taken you Government funding are researching new semiconductor equipment are using like research based tools some of these tools I've been guilty of this when I was in grad school like clued together like I've literally built together my tools like no way is that going to to when I say hey I got this to work and I go to Applied Materials or or any facilities like hey this is what happened they're like we got to start all over because we don't that's not a that's not a manufacturable process and that that is all too often is what happens and it's it's amusing and yet it's kind of sad because that means we start all over again and potentially like maybe the cool thing that discovered it it really needs to be it's not as cool anymore once you've go into manufacturing you know process because of some some process that happen so what we're trying to say here is that translating Research into actual you know Tech transfer capabilities requires that they're using we're talking the same language using the same tools and so we want to make sure that research have a facility and access to commercial grade tools and like a commercial grade process flow so that where they're actually not wasting time right not wasting developing a unique thing that's unique to their lab but developing something that actually can be easily transferable to a a a a a company go down down the line that would immensely help the solution transfer kind of Ip transfer kind of process and makes that that Innovation a lot more worthwhile efficiency it's more about making sure that's efficient what do you think about the sort of qualitative as well as quantitative difference is of how an implied Ventures or an Intel Capital think about spending money in this space as opposed to what incel does in the past they were actually they spent a lot of money actually on on tools and equipment and and things like that and my my my partner in crime here alen tangal who is one of the co-authors on the paper wrote she's our the investment partner that that I work with and we both do a lot of our micr Electronics Investments formerly she was at appli materials in the Venture office there and I would to what what she says and I a lot of and a lot of times that those organizations have traditionally been supportive if you look at the number of um these Strategic investment firms that are investing in Hardware have actually dwindled and and now there's only a handful of them and even some of those are very selective on what they're investing in and I think if you talk to Intel Capital guys they would say that they probably are are trying to move up the value chain and you know looking for more solutions right and rather than these point and Technologies I'm hopeful that with with relative to to Intel like with Pat Uh Ginger's announcement of their new strategy that that they're maintaining the the ability to actually manufacture as part of their kind of road map that the the understanding that there's a need for codesign right that um designing a chip and understanding How It's Made is important to actually making a really good product that mentality would go into this their idea of what investments we should be making and hopefully a lot of that will translate into an increase in in Hardware Investments but currently I think one of the challenges we're seeing again to motivate why we need a fund is the fact that there aren't many US investors in the space even the ones you think should be in the space the NCSA Bob work and Eric Schmidt got together wrote 800 pages about Ai and all its permutations and what the US government should do about it there was a microelectronics chapter which the main takeaway for me was trying to keep China at least two nodes behind the world's latest and greatest as well as other recommendations sorry I'm curious for your thoughts on that recommendation as well as others in that document sure yeah I think in general I thought that you thinking about what they're they're looking at AI but I thought they took a really good tact and broadened it into all the things that would impact and support kind of AI That's not just a software thing and that there is a large Hardware component a lot of the new algorithms are going to be really dependent on Harbor accelerators and you see a lot of money going into new Chips to actually accelerate some of these these new models and algorithms and that even with that the growth of algorithms in AI have outpa based Hardware development so there's an interesting open AI showed like a very interesting graph of the model complexity of these algorithms versus mors law you have a com at best 2x growth right in your performance of whereas I think the AI models are going like 7x growth and so no way are we going to be able to do the things that we are programming on traditional hardware and that we do need some new hardware platforms to be able to do that I think what there's what they're suggesting in terms of supporting and looking at new Harbor absolutely definitely true I think that the ideas of in terms of national us Centric view of you what we should be doing looking at new new architectures I think is very important looking at Advanced packaging which what they said is important and I think some of the things like 3D stacking which is we consider that as part of advanced packaging of how do we put CH chips together better not just on a plane but let's stack it up on top and use a third dimension to even get better performance there's a lot of options to doing that well I I I just want to say to our listeners who've made it this far please I hope you guys pause and Google all of the terms that Yen has has thrown over the course of this conversation because all of these things are just like very confusing and fascinating I would recommend YouTube actually as a way to as a first stop to get your head around these this new technology like seeing it in visual form of what 3D packaging actually means as someone who was a history major and took phys and took quantum physics for non-science majors in college this was a starting to wrap my head all around around all this is a real Challenge and I think I think you're seeing you're certainly seeing the kind of Growing Pains of this if you think it's hard for senators to understand how what Facebook's business model is the micro Electronics ecosystem and and supply chain is levels and levels more more complicated than understanding that Facebook sells ads I guess this is a question yeah you spent time on the hill what do you think is what advice do you have for you know the Senate staffer out there who is trying to be smart about supporting in an industry which is outrageously complicated and needs phds to to do a lot of I think the danger is you try to do a universal solution that you take it all in and I would rather see that offices really take what parts of this that are more relevant to like what their you know their bosses are champions of and and then using that as a way to support the this I movement maybe the right word that because I think there's because of how complex the whole system is right this there's a danger of trying to do all at once rather than what for instance for offices that are supportive of increasing Federal funding for research you areas I would say you should be different looking to is like new materials research let's start doing that because the current system that we have right now is all developed around silicon you know all the tools all facilities but by no means silic the best material there's other stomach conductors out there in fact if we were going to go into a EV future you know ones like gum nitr silicon carbide are are important ones to be looking into and let's make sure the US is is leading in develop of those material and oh by the way the US is a leading um manufacturer of tools and those tools will be essential to getting to operationalizing those materials working from a position with strength leveraging our our our ability to do that in these new material groups could keep our adversaries on their toes instead of saying hey we just spent billions of dollars on a silicon Fab ah got start over again and now make it silicon G nitri FB or something like that so I think so that's one area that I would if you were if you were to support like R&D research let's do that for the ones that are looking to like entrepreneurship and like competitiveness let's I would welcome you know support on strengthening that commercialization pipeline let's look at different ways that we're we're going to help entrepreneurs within Federal labs to you know get the resources they need to to to commercialize their their products and I think what we've been doing with emerg I think is one step towards that there's other players involved of course and finally I think ones that are looking at a larger strategy of how we should be dealing with foreign investments in this kind of age where it's not just the Government Federal funded research that's competing for ideas but Venture Capital private sector private equity and also from the Chinese kind of mixed use military civil Fusion we need to have our own weapons and I think support for a an investment fund is one of those key tools we should have in that toolbox and support for that so I think there's some offices that really can get hopefully get behind kind of those ideas I think I just also just want to to to say that in context of also this like recent anti-asian you know hate crimes and things like that we should also keep in context and and hold space for complexity here because because we're talking about a lot of times if it wasn't clear in this podcast we're talking about China A lot of times what they're doing and having is sometimes the language can be adversarial and somewhat hawkish and then we should be able to hold space for complexity yes they are competing with us in many different things we should be able to to fight with them on the idea the best ideas in order to to make sure that we get the future that we want and also punish them on things like what they're doing to with the Wagers and punishing of of them and holding them accountable for things that they're doing that are against kind of what we think is important for human rights at the same time recognize some of the language that are you know that conflu and anti-chinese sentiment does trickle down to a lot of the folks here many of the engineers and folks developing Technologies industry and that can result in this you know kind of situations that we're having right now so just want to make sure that folks hold that space for that complexity that we can be both we can recognize that they are adversaries in some aspects but we should also need to cooperate with them on climate change on global Health in M of this pandemic there's rooms for diplomacy and room to work together and as peoples we that with a you know large diaspora not just in China but larger Asian-American Community is get gets affected and we don't want that to get wrapped up in this whole hawkish anti-china sentiment so that is I think as I'm talking about technology I think is important to talk about how that affects the people and societies and both sides other the pond as well thanks for that Yan it's the the sort of vision like earlier in this conversation you put on your like as a technologist hat and it would be really great if more people could put on that technologist hat and see that there are common Global issues where technology is going to be useful and yes the US probably doesn't want we just saw new export controls on a on a company that's like making missile technology and clearly there's going to be a line with all this but there's also like a ton of global problems that technology is going to have like Chinese graduate students coming to the US and like having a good experience here and not feeling like they're going to walk out on the street and be beaten up the humanity will have a far better 20 Century if we're able to continue having the sort of academic and you know commercial exchanges that the the US and China have been able to to enjoy since since reform and opening and keeping that sort of perspective in mind that there are like other Global challenges that like sort of Technology can solve and it's not always about making sure that your supply chain is hardened from from China or whatever that's not always the the number one priority keeping that perspective is really important and I think a frame that folks should keep in mind as the sort of like new Cold War Dynamics seem to spin up left and right and we have China bills and politicians really seeing no downside from framing everything as like a let's be China context and