SPEAKER_00: it just strikes me that we're gonna have a future that's going to be a little bit more efficient than people think we're not going to boil all the oceans we're still going to have power for our appliances and we're going to have essentially infinite intelligence in our pockets and all of our screens and to me that's a pretty exciting future and so and we better we we will need to

SPEAKER_05: have infinite intelligence and that's the whole goal like if you ask nvidia if you ask us if you ask any other silicon company i think they're all trying to say that you know we're we're approaching it with some different approaches but basically it's like how can we bring like cost per token or cost per generation down so that we can afford to have infinite intelligence while

Chamath Palihapitiya: increasing our energy energy production as much as we can this week in startups is brought to you by

SPEAKER_08: alpha sense get deeper insights into your business with the power of ai search and market intelligence start with a free trial at alpha-sense.com twist vanta compliance and security shouldn't be a deal breaker for startups to win new business vanta makes it easy for companies to get a sock to report fast get one thousand dollars off for a limited time at vanta.com twist and oracle oracle cloud infrastructure or oci is a single platform for your infrastructure database application development and ai needs save up to 50 on your cloud bill at oracle.com twist hey everybody welcome back to this week in startups my name is alex and today i'm bringing you two startup interviews now both of these companies share a backer valor equity partners so if you want to know what one venture capital firm sees in the future well here is a good look at it first up we're going to talk to republic you may know them back in the day as an equity crowdfunding platform but since then it's gotten into quite a lot more activities a little bit of venture capital some secondary shares tokenization of assets you name it they're working on it i care about this company because it's democratizing finance and bringing more investment opportunities to more people so that way when my kids are older well they're going to have a lot more different things to put their capital into than i did when i was their age then we're going to talk to positron another company that wants to build chips for faster and more energy efficient ai inference compute given that around the world we're seeing data centers run into power availability problems what they are working on could have enormous application enormous revenues and also may just challenge nvidia a little bit around the edges in the next couple of years these chats were a lot of fun i learned a lot i hope you love them let's get into it here's

SPEAKER_13: republic set the stage now making normally out of reach assets and investing methods more accessible to everyday investors is big business robin hood grew on the back of zero cost trading and bringing access to more exotic financial trading tools to regular folks to pick an example now republic in contrast has long been known for its place in the crowdfunding market spun out of angel list the company has raised more than 200 million dollars including a well-known 150 million dollar round in late 2021 led by valor equity partners now in the intervening years republic has expanded its feature set far beyond traditional crowdfunding so to tell us more about the state of alternative investing and where republic sees our democratized financial world heading please welcome to the show it's kendrick when co-ceo of

SPEAKER_16: republic kendrick how you doing alex thank you so much for having me i'm doing fantastic i also love

SPEAKER_13: that we have in your background the and i say this with nothing but love the de facto silicon valley

SPEAKER_19: office it is standing desks multiple monitors and people typing and that is exactly my happy place i love to

SPEAKER_20: see it how many folks do you have behind you you know maybe about 40 or so and we're in silicon alley

SPEAKER_21: uh so new york has become quite uh the tech hub as well and uh we're super proud to be here but uh our heritage our route uh go all the way back to silicon valley a decade ago so this is off topic but

SPEAKER_13: i'm just curious while i have you here i'm talking about the geographic split because i'm on the east coast too uh as the co-ceo of a leading technology startup do you ever feel that magnetic drag back to san francisco because the way i hear it told in the ai era sf has once again regained its primacy as kind

SPEAKER_24: of if you will the new york city of tech you know when we launched uh republic back in 2016

SPEAKER_25: naval ravikon my uh mentor and boss at the time uh was like hey everyone is you know based in or come to silicon valley why are you looking to move to new york and my view is twofold one is that we already have uh angelis and in san francisco as the roots right that's how i get to know a lot of the vcs from my days at angel is and secondly if you're building fintech then i think it's a good thing to be based out of the financial capital of the world uh and that's obviously new york city so uh uh in the particular industry that we're in uh in the business that we're in i think that us being headquartered in new york with a lot of trips and events and uh and network uh back into the valley

SPEAKER_28: uh is is a very ideal way uh an ideal setup for us yeah i don't even want to know how many frequent

SPEAKER_29: flyer miles you've racked up flying to sfo over the years uh okay but let's get down to brass tacks

SPEAKER_13: here i know republic i've known about the company since it was founded i've always thought about it as a crowdfunding platform clearly you guys do a lot more and we'll talk about that in a second but just to get people caught up what is the state of crowdfunding today it was pitched as a way to really break open a lot of private companies for folks so has it lived up to that potential and is

SPEAKER_32: it still growing uh alex uh if i may just define crowdfunding first and foremost is that is so uh you know it's so broad kickstarter is crowdfunding indiegogo is crowdfunding angelis is crowdfunding where you have accredited uh investor coming in to co-invest in a deal uh republic only took advantage of a law that changed in 2016 that allow for anyone no matter what net worth what income can come in and back and invest and have equity upside in a wide a variety of businesses so fast

SPEAKER_25: forward nearly 10 years now there's no doubt that the business model has proven to work not just for early stage tech companies but late stage movie financing music financing crypto that is digital securities being fractionalized so i think that we're still in a long first inning this next era but there's no question that is well on the way but i think the true potential the size of it uh is something that is to be seen in in the the months and years ahead i appreciate the clarification i i probably

SPEAKER_29: should have said equity crowdfunding versus crowdfunding writ large i mean one of my favorite bands archspire is crowdfunding their next record that's not the same thing as what we're discussing

SPEAKER_13: here okay so fair enough but there was a change in late 2020 that raised the cap that a company could raise in equity crowdfunding dollar amount from i think it was uh kendrick 1.1 million to five and so i'm curious now four or five years past that point what impact did that have on an equity

SPEAKER_32: crowdfunding and also on republic's business you know it it brings uh later stage companies because

SPEAKER_25: one million dollars is still a lot of money but you know for a company is already in this series a series b series c five million dollars as a cap is a little bit more meaningful they can engage in bringing more customers more community members uh and you know i'm optimistic that cap will continue to go up uh you know to perhaps 20 millions and even there's been talk in in uh in dc of potentially expanding it to a hundred million dollars and uh and beyond so again just the very beginning but

SPEAKER_32: there's no question that currently as is we're working with large enterprises uh across a wide range of

SPEAKER_25: sectors and industries yeah who want to engage the public through this regulatory framework in the

SPEAKER_13: u.s so generally speaking i'm in favor of the cap going up uh i'm curious though is there a a ceiling that you would put on it is there a dollar amount that's probably too big or is there no real limit here in your view of how high we should raise that that that max limit you know i am definitely a firm

SPEAKER_32: believer in free market to do so sensibly but i think this arbitrary cap on five millions or ten

SPEAKER_25: millions uh i think it's a little bit artificial i think there are other uh ways to address uh to to protect investors interest rather than arbitrary threshold so if it were up to me i would not put a threshold on it similar to uh you know under regulation d or or if you're gonna go public uh you know for

SPEAKER_13: example yeah no i i'm with one that i was curious if there was any technical or uh hidden risk that i wasn't thinking of but i'm glad that we're aligned there now equity crowdfunding one part of the business clearly not the entire thing you also have republic capital which i believe has something like 500 600 million AUM is that a traditional venture capital vehicle that i should kind of think

SPEAKER_41: of uh as in competition with a sequoia or does it have a different posture on the market that i should

SPEAKER_25: keep in mind uh alex it's a great question and allow me to define what we are today and i think it ties everything in together the business of republic is building the infrastructure that can accommodate capital raising and community investing next generation what does it mean it means that the repertory framework to take in non-accredited retail the ability to accommodate institutional as well to spv or direct their ability to do secondary trading their ability to tokenize and put uh a

SPEAKER_32: whether it's a small business or a uh you know a certain crypto project on a revenue sharing basis on chain right so when we first launched the first piece was equity crowdfunding we added on capital and of course to generate revenue we did syndication and other things but think of it as different components that when all in together enable an enterprise to fractionalize tokenize and engage with the retail public globally with liquidity that's obviously a very ambitious plan and it took us 10 years to get to where we are today which is the first year that we have a completely functional infrastructure

SPEAKER_25: for rwa real war asset tokenization for true retail participation and we're the only one as far as i know in the market that have that complete regulatory framework now we do have you know we add on business line business model but no we're not looking to compete with sequoia we do spv we do syndication i'm wearing a shirt from one of our portfolio company called k2 uh but uh but the business acquires venture capital the business republic is in financial infrastructure and asset tokenization i don't want

SPEAKER_54: to ruin your day but have you taken a look at your cloud computing bill lately they probably give you some kind of deal to start but over time those bills start to add up well our friends over at oracle cloud infrastructure want to help you cut your cloud bill in half yes that's right 50 percent while you're getting better performance at the same time oci is a next generation cloud designed to work with any application including ai it's faster it's more secure and you can do it for less we're talking complete cloud infrastructure and services regardless of your specific setup or workload and oci costs significantly less than other clouds with a span of 50 interconnected cloud regions and more than 150 oci services apiece so you can access your cloud from anywhere and keep your prices consistently low worldwide so join modal skydance animation and more innovative ai tech companies who upgraded to oci and saved see if you qualify for half off at oracle.com twist that's oracle.com twist this offer is only for new us customers with a minimum commitment okay uh it's

SPEAKER_56: almost like you read my notes because that's where i was going to take us after a couple of questions so

SPEAKER_13: we're going to loop back to the tokenization and the platform element of this i just wanted to break down the different things you're doing today and then we'll talk about how they how they come together so republic capital then is a tool to collect uh retail and probably family office money and put it

SPEAKER_32: into spvs is that is that the main work that it does correct that that uh that you aggregate uh larger check side from family offices and institutions into spvs uh and deploy them into more mature

SPEAKER_25: companies this typically private equity of venture capital and that is republic capital and then you

SPEAKER_13: have a a separate product called republic venture that is only open to accredited investors so not the not the masses as it were but folks that meet a certain test or threshold set by the government and you and i have probably similar views about that uh but i'm curious can republic venture customers

SPEAKER_32: invest in the republic capital product uh yes i mean republic venture is just earlier stage uh

SPEAKER_25: and republic capital aims at larger check in later stage uh you know companies uh again so you can see within republic that we have businesses that almost replicate miniature version of the entire ecosystem right in the broader market you have venture firms you have pe firms you have platforms we have all of these things just so that we then can replicate um and change and bring forth the entire

SPEAKER_13: financial ecosystem and another element that you're bringing forward is secondary trading you guys purchased cedars which is now republic europe and you've also moved into secondary trading uh we talk a lot about secondary shares here on twist because you know there was the venture liquidity crisis everyone was desperate to get a little dpi i i'm curious kendrick how big of a business is the

SPEAKER_29: secondary trading part of republic today is it is it large is it small i'm just curious relative to

SPEAKER_25: the rest of what you do yes so today is still relatively small that is you mentioned cedars which is now republic europe that's our uk european side that does both primary and secondary trading for example if you go into republic europe you're gonna see revolut one of the uh largest uh private uh you know neobank out of europe you're gonna see shares of revolut being actively traded among non-accredited investor on republic europe in the us the regulatory framework is a lot more complicated that's why

SPEAKER_32: carter failed at it that's why forged and equity day and equities then only deal with you know high net worth investors that's why angel is there's some secondaries i mean i think we did the first one when i was still there and structured it but that's only for high net worth individuals no one has managed

SPEAKER_25: to do secondary trading for non-accredited investor in the united states a company that we are in the

SPEAKER_32: process of acquiring called iron axe and iron axe is a very unique set of licenses as an ats that has the

SPEAKER_25: ability to do so so we're very excited to roll out at scale true secondary trading of private securities for non-accredited investor sometime over the next six months or so kendrick when i hear the acronym ats

SPEAKER_68: i think applicant tracking system i presume that's not that's not what it means so can you define that

SPEAKER_21: acronym for us oh it's automated trading system so ah that makes way more sense notice you know an exchange i think this is just a legal thing a for for most people there's no difference but

SPEAKER_25: regulatory wise you call yourself an exchange you take on certain regulatory obligations if you call yourself an ats which is like an exchange light then uh then you're you know under a different regulatory

SPEAKER_29: framework so one i don't mean to be a spoil sport but one thing that has always made me a little bit leery of secondary investing is a lack of information and there seems to be a pretty big disparity at times between what primary investors get into venture round and what secondary investors may get later on if they purchase their shares on equity forge or on republic etc equity zen i guess or forge global how do you fix that how do you make sure there's enough information coming out from these companies to allow for a retail investor to take a position in revolute for example without just gambling yeah

SPEAKER_32: i alex is a great question first of all the same information uh disparity or the lack of or the

SPEAKER_25: insufficiency of applies to venture capital uh to the traditional private and public market i'll give you an example the uh investment on angel list right so if you have a doctor coming to angel list to look at a yc uh company uh yeah information available literally fits on one page and there's not really a ton more information there uh and of course we trust that that doctor can make the investment sensibly so about 10 years ago the sec and congress already made a decision that hey just because you're not a millionaire it doesn't mean that you shouldn't be able to participate in the private market in the same way that you know the information on amazon when you buy a product may be different than if you were to buy a product directly on the company's website but you got to trust in the market and in the public to make sensible decision for themselves now the law does add on a few things you know ats is in exchanges and broker dealers have the obligation to do their own due diligence and make sure that they introduce things that are suitable for their customers but i don't think that the public disclosure requirement for the public market to your ipo is some sort of a gold standard that lasts forever i don't know about you but i even though i'm relatively informed and do have a you know meaningful public equity portfolio i can't remember the last time i read a 10k or an ak or an sec filing

SPEAKER_32: before making an investment in tesla or facebook so at the end sure yeah but kendrick you're taking on

SPEAKER_13: so one i appreciate the clarity here this is very useful but it sounds like you as a company take on a lot of responsibility then to not bring trash essentially to your your customer so the due diligence is in

SPEAKER_82: in some ways uh mediated by republic uh in a way so our goal is to make sure that we have the legitimacy

SPEAKER_25: lens and of course we also apply an additional lens on suitability but yeah we don't aim nor is it

SPEAKER_32: feasible to say that these are high quality deals and if you invest right hand and you're going to make money online that's not how the industry works i do think that market maturation interface ahead is through education onboarding the investor base and then present legitimate that is non-fraudulent ideally companies that don't make misrepresentations and leave it to the general public to make that decision that if they have a thousand dollars to invest they should invest you know 15 or 20 dollars in 50 different companies and not putting a thousand dollars in one so it's about education and access not about some curative lens of any one particular firm or fund right you make sure

SPEAKER_29: it's a suitable investment and not a clear fraud but after that it's up to the people to make their own

SPEAKER_85: choices that's why that makes sense to me work faster isn't just my personal motto it's the key

SPEAKER_87: to success for busy founders everywhere startups have always had to move quickly but even with that pace that doesn't mean you can ignore security and regulatory requirements but fear not there's vanta their ai-powered platform makes it easy to ensure compliance in days and keeps you on top of everything so nothing holds you up today or down the road keep moving at startup speed with confidence because vanta scales with you with round the clock support in case you run into any problems they're going to be there for you fanta understands your company's needs because they're a fast-moving startup too in fact i'm an investor and man that company is growing because they have a great product and they care just last month vanta added a bunch of new features including ai-powered questionnaires that help you breeze through vendor security protocols and autonomous penetration testing baked directly into the platform twist listeners get a thousand dollars off by going to vanta.com twist that's v-a-n-t-a.com

SPEAKER_25: twist for a thousand dollars off there's a clear parallel to commerce in e-commerce as well right so like 40 years ago what people purchased which is basically what was produced and made

SPEAKER_32: locally with the advent of e-commerce amazon one-click purchase now we all buy things online with a magnitude of of volume and choices many things we don't even need and i think they've got to be the same lens that applies here which is when it comes to commerce or investing you make things accessible you do require the company to post fair information but you got to leave it to the general public to decide

SPEAKER_25: and participate in shaping the technological and the economy of the future that we're all going to be

SPEAKER_29: living in we could solve this by the way by just having greater disclosure requirements for private companies but that's my own hobby horse i won't bring that up today uh okay now kendrick we have to

SPEAKER_13: talk about a couple things here so there is republic note which is a quote revenue sharing digital security that allows you to benefit from the economic upside of select republic portfolio companies and you also have a tokenization business and a web3 console business and my question before you kind of brought this up earlier was going to be are you eventually going to have a single infrastructure layer that would underpin or undergird each kind of like pillar that we've described at the business because it seems like you're doing many things that all kind of point in the same

SPEAKER_32: direction is that where the company's going uh for sure uh alex it's already that way so okay uh from

SPEAKER_25: the technology and legal framework is one system that is the the you know let's say to to do for a company to

SPEAKER_32: be uh traded secondarily you have to have primary issuance when you deal with primary issue and you deal with some are non-accredited some are accredited some institutional some are non-us they require different regulatory accommodation and frameworks to bring them in so we over the years all of these things are part of the same operating system where in in the early days we add on distinct business lines only to take in revenue but the legal infrastructure and the technical infrastructure is one cohesive one that underpins a wide array of industry i again to use you know poorly amazon as an analogy once you have the e-payment and e-commerce infrastructure you can sell books you can sell

SPEAKER_25: grocery you can do movie streaming widely different sectors in exactly the same way it's just that when it comes to finance it takes a lot longer to build because it's much more regulated but we are building that e-finance infrastructure similar to amazon as an e-commerce infrastructure for that

SPEAKER_79: war kendrick would would the a better analogy be you're building the aws for investing like you're

SPEAKER_100: building a base level of like infrastructure to allow a variety of different things to happen on top

SPEAKER_32: of it yes uh no uh but there are different components right i think a when i think of aws i think just this is like storage uh in this one case when you talk about investing primary issue and secondary trading on a cross-border basis then you have to deal with banking uh disbursement all of these components that that that that that the power of wall street now is being redone pieces at a time by republic and a few other firms in the space so yes we're building one singular operating system but we're not elon so we have to generate revenue and not just you know for 10 years uh but but yes we still in the very early days of about to launch nothing that we've done today is meant to be you know the

SPEAKER_25: business republic the old data version of what we're about to launch ahead kendrick an absolute

SPEAKER_13: pleasure republic.com is the url and uh we'll have you back on in six months or a year to see how far things are going but uh i would not bet against you thanks for coming on the pleasure is all my

SPEAKER_15: thank you so much alex i keep really looking forward to next time all right so we're going

SPEAKER_00: to need enormous data centers all over the world the size of manhattan the size of why i mean they're going to eat all the power and take up all the space and drink up all the water or maybe not there are a couple companies out there that are working on different types of chips that are going to reduce the overall electricity demands of our future ai world i'm very excited about this because i'm a big ai bull and also i live in a state where we're trying to expand green power and it's pretty tough so to learn more about these chips and the companies behind them please join me in welcoming the ceo of positron mr matesh agarwal mateh how you doing i'm very good thank you so much for having me here

SPEAKER_109: i'm really excited to chat actually and i do have a comment about the power if you want oh i'll never

SPEAKER_110: take an offer i'll never not take an offer so go for it well one thing i want to clarify is as much

SPEAKER_109: as and we'll go into positron and how we are trying to be very power efficient but i

SPEAKER_05: personally am a very big believer in we should generate as much power as we can so there should

SPEAKER_109: be no restrictions on power generation just use that power that we're generating a lot better so

SPEAKER_112: that that's that's the comment so you you're an all the above kind of guy i presume small module

SPEAKER_00: reactors solar wind um okay for fun what's the most exotic form of power generation you favor in the next

SPEAKER_115: 10 years so i i i have one i i have one too uh electricity generation on waves uh like wave

SPEAKER_00: movements ocean waves uh tidal power essentially right now power exactly okay uh there's a company called exotic so that's well no no i appreciate that i think uh exo watt is a company that i've talked to and they're doing um storing industrial heat via lens to solar power and i was like oh wow

SPEAKER_109: that that's awesome okay yeah that's a bit more exhaust exotic the company that that popped in my head was pantalassa that is doing the tidal power uh thing where they're deploying data centers in the

SPEAKER_120: middle of ocean and using the electricity okay so free cooling essentially right i mean if you do

SPEAKER_00: data centers in space it's harder to have to radiate it and all that business but the oceans are big here's my concern though and we'll probably cut all this out of the episode but who cares uh if you put all the data centers in the ocean and we're dealing already with rising ocean temperatures

SPEAKER_123: are we are we just slowly boiling all of the remaining ice in the broader the effect is very

SPEAKER_109: very very minimal uh compared to what the heat is but yes that is actually a really good question

SPEAKER_125: i'm concerned because that will 100 come up in in their investor deck luckily i don't have to answer

SPEAKER_127: to that no not at all okay so positron let's talk about it so positron is a company that's building

SPEAKER_00: basically digital brains to run transformer architecture to power large language models as we understand them today but to make people care about that i think we need to talk about why gpus are not always the best computing choice for ai workloads and people will be surprised mitesh because nvidia is worth 80 trillion dollars today and everyone cannot get their hands on enough gpus so why are gpus not the be-all end-all for ai compute workloads yeah so the first thing

SPEAKER_129: i'll start with is gpus are like especially nvidia gpus are the major like over 90 percent deployment

SPEAKER_05: of compute today right across both training applications and inference applications so training is when you are you know making the model learn and inference when you're generating whatever you're generating tokens videos images right and today training is the majority of the compute spend so it's still well over 65 percent of the compute spend inferences the 30 35 remainder of the compute spend uh nvidia gpus super well defined and designed initially for shader processing back in like 10 years ago when ai was not as hot but then really found out that for matrix matrix multiplication which is similarly in that realm and which is exactly what is used for training application absolutely phenomenal you are generally flops bound so the amount of transistors and compute bound so the more you can push flops through the more you can really do more training workloads and much more efficiently and more more efficiently and and that's what nvidia gpus are phenomenal at training applications on inference side though inference is is a much more nuanced or more nuanced beast in the terms of it depends not only just on the flops but it depends on memory capacity and memory bandwidth and at any given point of time depending on how you're running the workloads any of those three could be your bottlenecks so nvidia is really good at inference but is it the most efficient at inference for certain types of workload no there there can be always you can always push the envelope just by existing silicon technology whether that's by maximizing memory bandwidth or memory capacity on that and that's where positron steps in this is where we are coming in and we're saying for inference application especially the frontier inference applications so really large like really in multi-trail parameter model sizes yeah video generations which require massive amounts of memory capacity and bandwidth we can effectively really scale out our memory capacity and memory bandwidth through our architecture and and because of that we can be very efficient at inference the last thing i'll end here where the reason why nvidia is not the most efficient at inference is because although nvidia is using the latest cutting edge high bandwidth memory what is called hbm in you know general terms uh for for for the memories thing which is the fastest memory bandwidth on theoretical specs that is available unfortunately when you run inference workloads on nvidia

SPEAKER_118: you are not able to fully utilize the available theoretical bandwidth if you want to run a successful company

SPEAKER_85: or maintain a profitable portfolio like i'm trying to do you need reliable data and without the most accurate and up-to-date information you're flying blind that's why you and your company need a partner like alpha sense they're the world's number one most trusted ai platform for market intelligence that means they can help you and your team track the latest economic trends before your competitors even know they exist hey listen here at twist we eat our own dog food we only pick advertisers and partners here on the show that we love their products alpha sense has a massive searchable database of public companies

SPEAKER_87: and that lets us quickly dig up the most relevant facts and context and their equity funding screener includes private companies not just the public ones they save us time and they make us super human so start your ai search and market intelligence journey today and get deeper insights to power your business we're even going to start you off with a free trial how great is that alpha-sense.com twist to get started make sure you use that url so they know we sent you and you can get this great deal that's where i

SPEAKER_127: wanted to go with this because i i did talk to the guys over at etched uh oh back in march and a really

SPEAKER_00: good chat learned a lot um i would say they're probably your closest competitor but they were really big on how nvidia gpus don't end up using all their compute power when running inference workloads they were like you know 30 40 efficient i presume it's the other bottlenecks that come into

SPEAKER_05: play there that you just mentioned yes correct exactly so because you're not able to because of the way the arc silicon and memory architecture is laid out and basically how the movement of bytes happen nvidia is not effectively really having the same ratio of memory uh movement to the amount of compute available so they have a lot more uh specked out on compute rather than the memory movement so they end up consuming less than what is efficient and there are many software tricks that all of us try to do whether that's spec decoding flash attention all those things it can drive up the needle but it goes up from 25 to 40 it doesn't which is still terrible yeah that's just super

SPEAKER_29: inefficient okay so let's get to what you guys have built to start which is the atlas system it has

SPEAKER_00: eight arch accelerator cards it has up to two terabytes of system memory it looks like a server

SPEAKER_05: rack tell me why it's awesome yeah so the main reason why we really wanted to do the atlas system was if you look at silicon companies uh you know any silicon company any any silicon company that over the last 10 years they've all gone down the route of okay we're going to design our asec or custom silicon and and it takes them three four five years to get a product out in the market what we did with atlas system was we chose fpga as our baseline product we knew that atlas has limitations in terms of specs like it only has 32 gigabyte of hpm memory or the number of flops on it is like 1 30th of h 100 gpu but what we knew is that if we put fpgs are programmable gate arrays like think of them as programmable gate array if you will exactly yeah so it's basically the the best way i can uh simply put is like a cf transistors right and you can shape how you want to structure the cf transistor however you want it so you can basically it's the versatility that's so important in fgps yeah yeah exactly you can inflict your your fpgas yeah onto your your inflict your hardware architecture on top of the fpga and and really show that as a proof of concept that your architecture is worth something to the world and it's you know it works basically so that's what we did with atlas is the point was it allowed us to get from the out of the blocks quickly get a product out in 18 months from from the launch and and get an atlas system out which is a typical for you server so this is a very standard uh for you system with eight gpus this is kind of been the standard so far until the gb 300s and 200s came about in in the ai accelerator world and you can stack a rack of them uh it's very very energy efficient you know each server is only two kilowatts for comparison an h100 gpu sorry h100 system is 10 kilowatts of power and what you can do is today with our atlas system and this is in production today we're shipping it out en masse in in in thousands of quantities basically you can run transformer architecture workload so all types of llms uh and what we've shown with our architecture is we can drive the same amount of memory bandwidth that we were talking at 35 40 percent for nvidia we can drive over 93 percent of available theoretical bandwidth uh and use that so even with fpgas which are as i said limited cards in terms of specs we can drive a a comparable performance to h100s and because we consume a lot less power and they are a lot less cheap that they're cheaper than h100s we can

Chamath Palihapitiya: really say that we are performance per dollar and performance per watt we are two and a half three three and a half x better than h100 depending on each of the inference workloads and if you're curious

SPEAKER_00: why he's saying performance per watt also it's because as we talked about at the top of the show we are often not just compute limited but electricity limited so both vectors really really matter yeah so i'm glad i'm glad you brought up the the asic versus um fpga point because i was curious what you started there i know your next system is going to be asic based but you said that it's a good proof of concept to show that your approach works so essentially the way that i'm reading them attention correct me if i'm wrong yeah is that you had an idea and you applied it via um fpgas to start because it's faster it kind of quick and dirty get it up show that it works and then you take the same overall concept map that to an asic takes longer but you'll have higher more performance technology down the road but using the same overall principles that you proved out with atlas and field

SPEAKER_129: programmable gatorades is that is that fair that is very very fair and and the big part there is you get

SPEAKER_05: immediate customer feedback so we have the systems deployed in production right now yes so people are testing their workloads on us we understand like so for example you know attend like when deep sea came out the attention mechanism changed from mha to nhla we got that real-time feedback how does that impact our architecture how our architecture performs for that performance rather than just doing some kind of emulation or simulation you have real world performance right feedback and and numbers that you get so that was a big part of of getting the product out quickly and then the second part is you know you talk about like a lot of we talk all of us hardware companies talk a lot about hardware the other big kind of feature is software you have to be really really good at it uh it inference the the the pressure of being cuda compatible or or being very easy is a lot lesser than on the training side on training side you cannot you know nvd absolutely like you know you cannot be not uh in the cura ecosystem on inference it's lesser so especially for bigger workloads but still you want to make it as easy where you know as long as you don't make people change their code that should be your baseline kind of target is that you don't want to make people change their code and this allows us to not only test that out real time we can literally import that same stack over on on to our asic because the fundamental elemental unit or the compute unit and the architecture stays the same what you're really boosting on the custom silicon are the massive amount of specs because you get to go to the latest process node you get to do real new memory architecture and technology and then you can really drive memory capacity and bandwidth up so that's kind of our goal and then lastly i'll say to just one point add it also helps with getting investments because you can really prove out your your chip works to to customers and and so on it also doesn't hurt that it drives a good chunk of

SPEAKER_29: revenue so yes atlas as a bridge over to asimov which will be your asic system that comes out uh i think

SPEAKER_00: you guys said 2026 yes how much does an atlas cost how many of them are you selling and are they mostly like proof of concepts with customers that want asimov and are waiting or are these people that really want to use atlas for what it is today to run inference workloads yeah so we we have now booked

SPEAKER_05: revenue also so when i said booked uh it means it has not converted into revenue because we have to produce and ship it to them but but we have now booked in tens of millions of dollars for atlas systems uh so as you said gives us real revenue and and the big part here is that the people that are buying today are are basically two major things one is they're buying because the system already performs and gives them enough of uh of a performance uh leverage for their use cases that they're thinking okay this is worthwhile investment to get my return on capital in you know 18 24 months so that's kind of the time period and then the second thing is you know people want optionality people at least want to give a chance to non-nvidia accelerators and i think this is where they're like okay this also serves as a test for your company can you productionize the system first thing second is your architecture worth anything like you know does it actually work and and does it actually scale out and the way you predict and simulate and it and and it really drives so what we call like maybe we're using the term maybe we are using the term wrongly but like we call it the wrong socket strategy it's like hey you you you buy atlas uh you you pay for you know we make money we make revenue we make profits and and most importantly with atlas we're not saying that we're going to scale out revenue to hundreds of millions or billions right i think it's it's it's unfair to expect that off of the atlas system but what we really want is the customer list that will spend that much money on the azimov and titan system and that's kind of how we really approach it so we call it the wrong socket

SPEAKER_00: kind of uh approach uh there well no i i really appreciate it it's always good to see a company that has a vision not just for their first product but their second and third i mean i mostly prepped on atlas and azimov titan i was like that's gonna come later we'll talk about that next year um but i i appreciate that now the customer mix is interesting here because i'm curious if you guys are selling these into the the hyperscalers or if your early customer base for atlas is more like i don't know companies that want to stand up their own inference stack because they're tired of paying someone else's

SPEAKER_05: margin on their compute needs yeah so so the two public companies that we have so the two companies that we publicly announced are cloudflare and parasel uh so cloudflare is a content distribution network parasel is an inference as a service provider right so so kind of different use cases but they both are looking at it from uh cost effectiveness and with cloudflare power effectiveness because they have these data centers in metropolitan cities they can't you know supply more power to it or can't liquid cool it right so they need an air cool setup how to drive more tokens from a given amount of power so that's kind of where they get interested in terms of other customer bases that we have not announced uh it is a mix of a neo cloud a hyperscaler to your point and and you know we'll we'll announce that in due time and in coming months uh so we get to it lambda or a core weave and then

SPEAKER_41: in aws or an azure just to put some names to the categories that you're describing i'm not saying there's the actual customers correct i'm just trying to explain to people yes yeah yeah to the categories

Chamath Palihapitiya: and then uh also uh we're now starting to get some traction in the financial trading uh ecosystem

SPEAKER_129: right so that makes a lot of sense to me yeah so those are the the ecosystems that we are really

SPEAKER_05: going after and again the goal is to to get them really excited about uh what it represents for asimov because where we are going with asimov is somewhere no other chip will be like so if you draw on a map like you know okay memory capacity memory bandwidth and uh uh and uh the computability the big uh claim to fame for asimov for us is it's going to have two terabytes of memory capacity so just to give you a comparison as i was launching end of 2026 nvidia will be on the rubin generation which is the one after blackwell uh so that's the next generation that they will be around starting to ship it out then rubin initially will have 384 gigabytes so that's 0.38 terabytes of of memory

SPEAKER_04: will have two terabytes so you're looking at a 5x uh memory capacity differential between nvidia rubin

SPEAKER_127: and and positron as it is okay at the top of the chat you said currently ai workloads are like 65

SPEAKER_03: train 35 inference yes i presume that was 90 10 a few years back yeah a couple of years back yeah so how

SPEAKER_129: long until it's 50 50. uh so what i do know is what that rate of inference spend is growing so this

SPEAKER_05: year roughly 105 billion dollars of compute spend like not data center just the compute spend that goes in an inference data center uh that will be spent on inference 105 billion it's estimated that it'll reach 300 billion so 250 billion in 2027 and roughly 350 to 400 billion in 2028 so that's kind of where it'll grow to the inference compute spend the training spend is already well over uh you know 300 350 billion today uh and and the thing is if xai keeps on increasing the colossus if openai keeps on actually spending the stargate uh full full-on project anthropic spends with tranium i actually see the tranium as a training spend go up about a 500 billion to 600 billion mark by 2028 so you're looking at a trillion compute spend and on training so it might maintain that ratio of 60 40 for the foreseeable future or in that kind of uh territory but there will be an inflection point someday i don't know whether it's three years five years ten years but but it will have that that's very interesting because

SPEAKER_29: i listening to you talk i was like why isn't nvidia trying you know getting off of their gpu train and

SPEAKER_00: having a second effort to just focus on inference because what you guys are building and what etch is working on and a couple of other companies really seems to me like the right approach to handling inference compute because you don't need to have the same stack that you use for training precisely but if if the training market is going to stay the majority case for ai compute for the next couple of years never mind i take it all back nvidia is barking up the right tree uh but you know in 10 years though we're not gonna i presume we'll spend a lot more on ai inference than ai training and so i think that that long term you guys are going to end up with the majority of the market it just may be a

SPEAKER_129: little further out than i thought i guess well well yeah and like look two things i think i'm going to be

SPEAKER_05: grounded about this and actually not even humble just for the sake of being i'm going to be very humble about it nvidia not only is the most valuable company in the world they are one of the smartest companies in the world it's not like they do not understand this differential between where the training workloads and inference workloads and efficiencies are right if what they're saying is there is no option in the market today that can even compete with the gpus on inference like and they're absolutely spot on right like that is true like that that is actually a true statement today and they're saying so why divert our margin stack where they make 75 margins on this this very high end cards to like you know to create cheaper or or more efficient interest i'm not saying that they're not trying to make it more efficient and then the second part which is where the humbleness comes in is the proof is in the pudding none of us or esht or any other company for that matter literally zero other companies have proven out a better cost efficiency setup than nvidia today in the market right like we can we can claim for for atlas for example we can claim it for certain transformer models or certain types of transfer models but if you look at the overall market no one has come out and said yeah like oh everything you know we can beat nvidia across the board because no one can and and even for the foreseeable future that's not going to be the case it's about finding the right niches like for example for asimov two terabyte of memory what does that immediately get us you can fit a frontier and model so like grok 4 is expected to be 2.4 trillion that's what the rumor mill suggests you can fit that on a single chip of our card of asimov whereas you need like four or five gpus to fit that weights of the model or video generation today video generation is limited to like 8 second or 10 second or 20 second clips with that much memory now you can generate a continuous video to a minute two minute you know even multiple minutes long so so you have to find your use cases where you stand out against nvidia and go for that and this is also why nvidia wouldn't be like hey we're going to create a single chip for every application because why are they going to sacrifice the margins for for

SPEAKER_29: for their but the cool thing is because nvidia is going after what is currently the majority case and the high margin case and the proven out case and the no right they're gonna make a lot of

SPEAKER_00: money but they're gonna leave some gaps that you can step absolutely yes so i want to ask about that because uh for fun i looked up in videos r d spend for the last quarter and it was 4.3 billion dollars uh your last round was about 50 million so about like i don't know 27 minutes of nvidia's r d spend uh i know companies like valor are backing you but i'm curious are you guys going to need an order of magnitude more money to pursue the asimov vision and then putting them together to make titan or is

SPEAKER_41: it going to be more capital efficient than nvidia's r d spend might lead me to think to get you through

SPEAKER_05: your 2026 2027 ish roadmap yeah two things on that first an internal kind of introspection we are very capital efficient so we got our atlas out with only few single digit millions of money raised and uh and with asimov actually the and look valor is is a well-known fund for their diligence they looked at our kind of capital spending plan they like capital efficient companies uh and uh you know when they looked at us and they saw that with the series a we could tape out our asimov chip so you are right generally or traditionally if you look at silicon companies they have raised 300 400 million dollars to tape out a single chip we are going into the market with a total raise so far of 75 million dollars including our seed and series a and we are saying that we're going to tape out uh our first generation ship in that 75 to 80 million dollar uh all right range and then to your point after that when it comes to production scaling it doesn't mean that we are done raising i i would be lying to you if i sit here that's saying that hey like we're not planning on raising more we absolutely will raise more uh but i think that will come for scaling off our asimov into titan systems as as you said and and really going out to market but you know we don't like personally we have a very strong person both my strong belief and and and the company we have a very strong belief that you know you're only showing market validity and valuableness if you are both uh doing that at a with a very huge amount of capital efficiency and all and your chip is only worth it if the market wants to pay for it what i mean by that is what we really want is third parties and you know hyperscalers and cloud providers and companies buying our chips we what we don't want to do is like just like hey build out our own cloud and then sell token as a service right because that's kind of where the economics values like whether the companies are economically profitable or not is not clear so we really like we want to do what nvidia amd does which is sell to third parties uh really and then that's kind of where the vision goes and if you're selling to third parties which means you're making revenue which means that profit can uh fund your future tape outs uh so that's the introspection side of things with nvidia's yes uh with nvidia's 4.3 billion number sorry just quickly addressing that obviously they're not just spending that on on their silicon the nvidia nvidia is a company that is just even you know it's not just a silicon company i know that 200 billion dollars worth of their revenue is driven by silicon but they're spending r&d on robotics on networking actually one of the key things self-driving with their

SPEAKER_04: thor system etc yeah yeah and and like one of the things that i was like i genuinely think where

SPEAKER_05: their r&d expense has has been really worth it is looking at their at least projection of their optical networking and and what they have the cpo the the the that they have projected at the gtc it was really phenomenal to see what they're uh kind of scratching on the surface there right so yeah so they spend money on a lot of things so actually again i come back to it i'm i'm very both obviously like i'm very bullish about positron what it can achieve from market share perspective because the market is like going to be 300 billion you know we're going to achieve a big part of that inference market obviously that's the goal that's what we're trying for uh but i also see kind of where nvidia spending rnd on different areas of domains of market it's not just training it's not just

Chamath Palihapitiya: inference you know you talk to a self-driving robotics other things as well i guess what i'm trying

SPEAKER_29: to drive at is like is there enough private capital in the market that is available to companies like

SPEAKER_00: positron to do the work you need to do to earn your spot in one of these niches that we're describing like does does valor and friends have enough gumption to give you the capital you need when you need it because we there's been a little bit of a sentiment shift about ai in the last couple of months i think gpt5 i like gpt5 me too i love that less impressed uh and so it makes me wonder about like how how open the wallets are as we look into 2026. yeah i think uh the interesting thing is

SPEAKER_129: like you know obviously if if i speak about valor or like i have to give them kudos and it's the same thing with valor and atreides you know uh antonio gavin and and these guys they backed unfashionable

SPEAKER_05: thesis early with us right you know like you know you know it the fashionable thing is obviously nvidia is going to keep on being the 94 of the market or 92 of the market but even if they are the market is growing so rapidly that the rest of the market is still pretty large uh but the second thing that i will say is the answer is absolutely unequivocally yes like you know with even just like you know like if you look at valor and atreides and and dfj who are backers for series a generally these are growth funds right like they they generally back companies in series b series c like when you're raising growth rounds they came in early in depositron uh and and that's not because you know uh they just fashion one day that we want to be early stage investors i think it's because they basically decided that hey this company if it actually proves and delivers and executes on what they're showing and promising is actually going to be on a very fast growth curve and they're going to be there to back them up because again that's where their actually main uh capital comes in is is like especially valor if you if you look at antonio's thesis right he he doubles down every single time on a company once they show uh execution and and success so so that's that's that's we have no doubts about that uh both on our ability to to prove it out to the market and and that there's capital available for us to keep on growing and i'm really finding a very very good uh revenue stream for us

SPEAKER_29: okay one last question before i let you go and it's just about the the durability of transformer

SPEAKER_00: based ai like you guys clearly have made a directional bet on transformers staying at the forefront of ai progress and as i alluded to a second ago some people are a little bit worried about uh training walls or you know token prices not coming down as we burn more than reasoning models and there was that mit study that everyone i think kind of misread but took to the fences so tell me a little bit about your just your and pelzadron's confidence in transformers staying at the heart of ai and hopefully as well that there's still good performance gains to be had out of transformer

SPEAKER_05: based llms in the future yeah can i say one thing to to begin with in terms of clarification though like we are fundamentally a linear algebra accelerator right what we are accelerating is the matrix multiplication and and so and you know if you if you look at back 30 years ago and i think if you look forward 30 years ago the fundamentals of deep learning will be in the matrix multiplication so linear algebra will still be the driver of the different whether it's transformer architecture but there was cnn and whether it will be what all of the future architecture of the world is like i

SPEAKER_03: would attack yeah are you working towards saying that i am being overly specific by saying transformer but really what i'm talking about is just matrix multiplication

SPEAKER_05: you are being i'm saying you are being overly specific that our our marketing word is yes we focus on transforms because that's okay the use cases are and that's where we actually really stand out as well but if tomorrow something changes like you know a new model architecture comes out for us we are not actually specifically tied to transformer what we can do we we are you know just like gpus can adopt a new architecture we absolutely can adopt to the new architecture it'll just be we have to we'll have to spend more time on instruction sets and then you know software that's a

SPEAKER_04: software point not a hardware point correct yeah from from a from an architectural perspective we can

SPEAKER_05: actually serve much more wider like we even today with our asimov and with our fpgs we can serve diffusion we can serve other types of models it's we have focused on transformer because that's where you know over 90 of generation or inference use cases today lies and also that's where people are really deploying it so from our side really we're making some decisions to optimize for it like for example transform big transformer models require lots of memory so we are going and chasing after

SPEAKER_198: having a lot of memory on our chip rather than so you can run those two pro two trillion models for

SPEAKER_05: yeah exactly on a single chip exactly right but but from an architectural perspective we can actually

SPEAKER_00: run much more broader ecosystem of models for inference not training well um normally i i end these chats with uh well thanks for coming on where can people find you online and what's the job you're hiring for

SPEAKER_200: that you're having a hard time filling but instead i'm going to ask you this um can i give you five

SPEAKER_29: dollars in your next round absolutely yes to be clear i'm kidding i'm just i'm very impressed and i'm also i'm also very i'm still a big ai bull yeah and i think that when i think about the asimov's

SPEAKER_127: you know system being put into like groups of eight and solace titans it's all it just strikes me that

SPEAKER_00: we're going to have a future that's going to be a little bit more efficient than people think we're not going to boil all the oceans we're still going to have power for our appliances and we're going to have essentially infinite intelligence in our pockets and all of our screens and to me that's a

SPEAKER_05: pretty exciting future so and we better we will need to have infinite intelligence and that's the whole goal like if you ask nvidia if you ask us if you ask any other silicon company i think they're all trying to say that you know we are we are approaching it with some different approaches but basically it's like how can we bring like cost per token or cost per generation down so that we can afford to have

Chamath Palihapitiya: infinite intelligence while increasing our energy energy production as much as we can well i'm totally

SPEAKER_00: here for i appreciate it so much it's uh positron.ai if people want to go take a look at it and it's please um when asimov comes out will you come back on and tell us how it's going i absolutely i would be

SPEAKER_05: i would be the happiest person to come back on because that would actually put us in to an untouched territory as i said in terms of a couple of those specs that you're targeting there late 2026 you

SPEAKER_04: said right correct we are taping out q3 end of 2026 and then uh production so initial sample

Chamath Palihapitiya: systems will be late 2026 and production scaling in early 27 that's kind of the schedule well that

SPEAKER_127: means you have quite a lot of homework to do i'll see you in 365 days thank you for coming on man

SPEAKER_107: my pleasure thank you so much for having me