The Solana Bet | Anatoly Yakovenko

Transcript

like the gas Wars that occur on ethereum are extracting value from the use cases that are running on it from the users from the Liquidators from the actual math Bots if you think MAV isn't that positive the gas is actually taking away from it and it is giving it back to the to the ethereum holders but I think over the long term there's no way that level of extraction survives Mass adoption it's just not how technology works alrighty everyone welcome back to another episode of zero X research today we are recording on April 24th and we have a great interview for you this time with Anatoly the co-founder of Solana we

know we've been on a huge modular thesis binge in the past couple episodes we really wanted someone to come on and give us the counter argument why is all this wrong what is the what is the other solution and why is that one right and no one else better to do this than Anatoly the founder of a monolithic blockchain that's really based around high performance High through foot high throughput low fees um just a totally different approach than really the modular thesis of specializing each blockchain for one individual component but as always we are joined by two of the blockworks research analysts to kind of give us a market recap of everything we've had going on since the last episode was released uh this time it's Ren in Xerox pibbles uh Rand I'll throw it to you first man who do you got in the hot seat

or cool Throne this week yeah sure I have a hot seat this week this week my hot tedis Mev blocker so if you're paying attention probably around two weeks ago Cal stop and a few other D5 protocols and a few other Mev infrastructure protocols work together to release Mev blocker for those that who don't know Mev blocker is an RPC that basically prevents your trades on the decks more often than not from getting Fun Run or a sandwich and I think this week a few people have learned how the RPC Works in itself and how it prevents the front running and there's been several points of contention within the Mev convenient community so

specifically with me V blocker a Searcher can see the content of a transaction without the signature so that does mean that if you are a researcher and if a user is going to buy a token they can still try to Fun Run you just perhaps not atomically in a separate bundle right so in order to combat this what Mev blocking does is that it sends a number of fake transactions to Searchers within the public member so if you're a user wanting to buy token X mov blocker basically can generate two more transactions um that basically say that you are going to buy token Y and you're going to buy token Z instead so now basically if you're a Searcher and if you want to buy

token Y which the user is not actually buying you're gonna lose money there because there's no one to sort of like close your trade out you're just gonna buy the token and you're gonna sell to yourself and you lose money on fees or for example like the Dex fees and so because of that the Searcher doesn't know what the X Y or Z is real and that's essentially the mechanism in which Mev blocker quote unquote guarantees that your transaction won't be fun run or sandwich and as far as for me that just feels like a weird mechanism to you in terms of sure you want to share information about your transaction but you don't want to share too much to the point where a Searcher can Fun Run you and then your solution becomes spamming

the mempool with random fake transactions and it just seems inefficient for lack of a better term yeah it does seem like an interesting way to kind of go about this problem um but what's interesting is it I I agree with you it feels like rudimentary in a way but if it works and that's our best current solution then does it matter that it feels like overly simplistic I don't know crypto we always bring up like razors and whatnot so like Occam's razor maybe the simplest way is the best way yeah and also I've been using maybe blocker and it's really fast I remember when flashbots dropped there Mev mitigation our PC I feel like the transactions were a little slower um but me V blockers really fast I like

it that actually feeds pretty well into my hot seat for the week which is just the General market we saw a ton of meme coins get launched everyone was trading Pepe LSD Doge um wojak neon or whatever it was called like it just got a little bit wild I think we all should have known there's a little toppy in the market but yeah Mev extractors had an absolute field day uh with all the meme coin action going on and it was kind of interesting to me because I noticed uh that the L2 Dex volume is kind of on a rise or on the rise but then I looked at Uni V3 on arbitrum and there's really not that much volume on Meme coins but then you go to UNI V2 on ethereum and the top 10 pairs are pretty much all meme coins

Pepe being number one with like 400 million dollars of seven day volume so I'm not really sure what the reasoning is maybe it's because uni V2 isn't really on many l2s but those l2s aren't extracting Mev and then they have lower transaction fee so I don't understand why meme season wouldn't take place on an L2 maybe it's just better marketing more people paying attention to like the hot contracts on ethereum mainnet not really sure uh but yeah nonetheless uh I've got the market in General on the hot seat definitely definitely got a little complacent over the over the past couple weeks I agree with you it really feels like we we should have seen this one coming when last week all the slack channels popping off with these different meme coins but um yeah nonetheless

it did get frothy I think we actually put ARB sellers in the hot seat last week which maybe we need to be in the hot seat for for throwing that one there because uh arba has been one of the biggest losers just being a higher beta coin over the last couple weeks uh those tend to indeed be the biggest losers but um yeah I think the point about meme coins is interesting right and it I don't know it brings up my inner conspiracy theorists like is there a cabal that's launching these meme tokens that actually see traction and tend to have like similar wallet histories linking a couple of them together uh and then you start to think all right well this is all happening on eth like it's a great way to say you know there's still eth burn it's still proving things to be

you know like proving like economic activity because on the surface you just see Dex volumes are high and it does look like that but I don't know I haven't dug in dug into that theory at all so it's really just a baseless claim but I don't know there's also the idea I guess that if you're going to launch a meme token you want to kind of go to where the wealth is rather than trying to pull the wealth somewhere else and by far the most wealth is on ethereum so it's like make it easy for people to buy don't make me Bridge anywhere else just throw the meme token here yeah definitely and a fun fact for all the listeners is that half of our team got wrecked on neon coin last week so if the dev is listening to this I wish you a very bad day

um but we started off strong in wojak early in the quarter Ren you had a cool Theory too around uh like kind of like Mev Bots like purposefully I guess putting these meme coins out there to try and get attention and extract map do you think there's much legitimacy to that theory I think that Varsity or at least that's a conspiracy theory I believe in because on one hand most of the AUM within crypto slides on ethereum right rather than I'll become an optimism but perhaps you could also argue that if it did launch an option of optimism it would have just seen a lot more volume just because transaction fees are lower but also you can't really there's no Mev for you to extract on the

original optimism today given those centralized sequences so I think I don't know I I like to believe it um but also like just some interesting stats right like Jared from Subway the famous Mev bot last week made around two million dollars in profits which is a crazy amount he was basically out of the trade volume within the Pepe the war Jack or the mean Point improves the guy process I think 750 million in volume between like if and all the main points which is pretty ridiculous for like any researcher and I think a few other notable people within the Meb Community have also pointed out that there was a combination of factors that made it so lucrative for Mev

participants last week right one was that apparently there were Bots that were designed to make any user cell transaction fail right so they were basically just massive other transactions front running and basically make the snippet within the Poo higher than your suffrage tolerance and to basically give the impression to users that your transaction is going to fail unless you set a really high slippage tolerance and I can't confirm it I haven't looked at the data but that's apparently what's happened there were votes like in those liquidity pools intentionally like messing with prices and making everyone's trades fail and so everyone ended up putting really really ridiculous like slippages and if you went to look at for example eigenfai the

war Jack the PV and if we looked at sort of the top profiting Mev transactions some of them were really ridiculous like the MVP profited like 30k and the end user got like 500 back from a chance that I can switch implies a stiffer tolerance of like 99 which I'm not sure why would you ever do that's I guess a summary of our state of the market right the market moves against us and we simply bring out the conspiracy theories fibbles who you got on the uh the cool drone this week yeah so I have the Mad Lads slash xnft backpack team it was a super fun mint that was my first mint on Solana in quite a while and uh it was

super cool like Thursday night when they were supposed to men it got pushed back an hour then another hour then a whole day uh because they're getting ddosed I think they had upwards of like 2 billion requests at one point um but they had a successful mint on Friday night and it's really neat because it's like the first public xnft mint um and that's like an executable nft so you can have like xnfts that are adapts that are live in your own wallet so this was the first mint that was live in the backpack wallet it was a super cool experience and overall I just really like the backpack quality yeah we've been harping on these for the last week and a half two weeks and it was fun to

actually experience the mint firsthand I was actually going to be really pissed because I it was supposed to be at seven o'clock on Friday after they delayed it today and then they did 10 minutes and then 10 more minutes and I was like oh my God this is about to be a repeat of last night and it's like a Friday night at 7 20 I'm like this is the last thing I feel like doing but yeah super excited about these things um I scooped one as well I know the floor has been kind of a rocket ship it's Monday April 24th but the floor is sitting you know mid high 70s so I'll be curious to see if the price action actually sustained remains I saw some people on Twitter talking about how Haiti swap which is like a pseudo swap ripoff on Solana um acquired like 225 mad Lads in order

to kind of have some protocol and liquidity and drive some volume over there so apparently that's a bad thing for the longevity of the floor price but I don't really know I don't stick around Solana too much uh during my like personal use of John chain so not not super sure yeah some cooler things to point out about the Mad Lads mint is one the Solana Network didn't go down so it's really cool I mean they get tons of flack for that you know it's happened a few times but I mean there's quite a bit of demand and things worked out fine uh another thing the flag is that there's going to be some sort of snapshot on April 29th I think and we have no idea what that's for

on your first point there we kind of get into this with the talk on uh with Anatoly but briefly you know the isolated fee markets and some of the other technological improvements that Solana's been working on over the last six to 12 months have really come to fruition in this moment right massive demand it actually crashed the backpack app itself but yet the activity on chain you wouldn't even be able to tell right so we'll get into the into the interview of what the uh isolated few markets do but basically uh with a global fee Market if one piece of the ecosystem is getting really expensive to use a lot of demand for it then everybody pays more fees we see that on ethereum all right if there's an nft mint then all of a sudden selling it making uni swap transaction

becomes 5x more expensive but in the case of Solana and their isolated fee markets they're just because there's one active mint not other users don't have to pay more and that's something that I wouldn't be surprised to see become more of like a crypto standard if we fast forward a couple years from now that just seems like a better way to build a blockchain yeah going back to xnfts I think it introduces some sort of unique mental models for people to think about right for example in the past if you wanted to gate access to a protocol or quote-unquote and app a very common method was to issue a limited edition nft then you would have to hold that nft in your wallet then you could access the protocol right but now with X nfts

rather than issuing limited edition nfts to access a protocol you could issue limited edition apps or the X nfts themselves and I think it's just a very interesting way of thinking about the possibilities with executable executable code within an nft and I also asked some guy at Solana I think some person on the developer relationship team in terms of what's possible Right for example if I have to update the X nft do I have to reissue a new x x nft and that's just no you know I thought it was yes but apparently all you need to do is provide an updated bundle and we upload it then the xnft will update its metadata and for example you're also able to have

multiplayer games work on xnfts which is also I don't think that was the case because all the games within the backpack app were a single player you know like 2048 Flappy Bird I was like it was like really like old school fun to play games but apparently there's a way to sort of make an online game which validates the state and basically stores a state hash and so you can validate like the state of the game and multiple people can play Within the same xnft so I just thought like those are very interesting to think about yeah that is super cool and plus like just think about the security upgrade for users like if it's in the App Store and it's trending it's probably pretty safe um versus I still triple check addresses

when I type them in manually if I don't have them bookmarked like I'll go to the docs and then I'll go to like their Twitter and make sure that I have the right URL before I connect so it'd be really nice to just have that all in one spot within your own wallet but Dan I know you uh got a cool Throne this week what you got yeah that's right I'll uh take us back to ethereum land and I'm actually put in one of the OG defy apps on the cool Throne curve Finance so I feel like we've been harping on their stable coin that's been like such a long time coming um but over the weekend we saw some interesting activity around Xerox babe one of the curved deployer addresses uh maybe we can link that address in the show notes just uh if anybody wants a

free little Alpha to follow that's a great address to throw on the etherscan uh watch list but nonetheless this address had some interesting activity around launching and testing new contracts and some of these were appeared to be stable swap pools with Ethan native as a as the pairing um and if we check the GitHub there was a couple updates around like test deployment of a dummy contract and things like test new price Oracle and one of the things that they've been working on is creating a price Oracle that is an alternative to use to chain link and so this kind of starts to feel like it's the last piece before we see curve USD the stable coin lending

protocol go live because obviously with a lending protocol you have collateral and you need to know the price of that collateral to make these you know pendulum type liquidations and this is going to be a huge unlock for a lot of different reasons aside from the obvious ones like vertical integration for the protocol right so also adding a stable coin and a lending protocol on top of the decentralized exchange it also gets the you know protocol Revenue diversification benefit of that as well but to me there's a more exciting unlock here and that's really like creating this Fearless transfer market between different types of listed collaterals and so what I mean by that this is actually first flagged me by curve cap he's got a great sub stack you should definitely check that out but he kind of came brought this idea to me

where if you have so assume there's two collateral types rapid for example each of those would have its own amm so its own pool and they'd have like liquidity within this pool well if you assume that this is the case then these pools are natively making generating fee revenue from interest freeze and which means that the protocol can set the trading freeze of these two pools to near zero or let's just assume zero and it would still be generating Revenue even if uh swaps are taking running through this pool uh at a zero percent trading fee so basically curve USD can come like this Fearless link between different types of collateral right because that's in the world where we only have two collaterals but let's

assume there's now 10 or a thousand collaterals well this stable coin then becomes like this Fearless transfer mechanism and the more depth the larger trades you can push through these pools and it kind of just like makes an interesting scenario where curve is a little bit more prepared for this race to the bottom and fees that we kind of expect to see in The decentralized Exchange Market so it's just kind of like another interesting element for how this stable coin lending protocol gets a little more interesting than just the on the surface like oh we're expanding what the protocol does but this also comes at time when we see more developments being pushed out by the curve team right so the first leg of the try crypto upgraded contracts was just rolled out as well earlier today on the 24th and so the try

crypto upgraded contracts are a huge unlock and honestly to me this is even more exciting than the stable coin um for a very simple reason so uh the try crypto pool is one of the largest pools on curve this pool holds ethereum um wrap Bitcoin and usdt tether and the ability to it's like using the curve V2 amm which is a concentrated liquidity amm that automatically adjusts pools liquidity around a central point so it's kind of like in contrast to UNI swap V3 which allows LPS to move their liquidity across different ranges selected individually by each LP in a curved V2 pool the point of concentration is automatically updated based on like an

exponential moving average of the swaps that are run through the pool so when the exponential moving average moves too far away from the current market price that is when the next Trader will be tasked with running a computation that rebalances the pool's liquidity now the issue is that that computation is pretty expensive and you're putting it on the bird like the burden falls on the trader which Traders are the ones generating your fees so it's not that great that you're kind of putting this burden on them because ultimately the gas costs to execute swaps in that pool will increase well the upgraded version of the try crypto contracts dramatically reduces this cost the early tests were around the 4X Improvement and if that's the case then when it gets pushed out to

mainnet which it feels like it's hours or if not at most days at this point um we'll really see if that 4X Improvement was realized but what this will do if that Forex Improvement is true then we will have the an average swap will be cheaper on curve than it is on uni Swap and a lot of the time it looks like the curved V2 model can throw you a better price than uni swap V3 could the issue was we really missed out that on um the gas fee difference right so it's more expensive to run this trade than the savings you'd get on a better price was removed by the extra gas cost that you would have to pay to execute and lastly one of the biggest issues that curve is facing right now

um in terms of like achieving the most Dex volume possible is the absence of an Ethan usdc pairing right now that pair accounts for about 42 43 percent of uni swap's volume on each mainnet and curve just has no exposure to that to run an ethi to USD swap using only curve you'd have to be routed through two different pools and again it comes down to gas costs that's just an expensive of trade so uni swap is just just absolutely eating all that volume and it'll be very interesting when you see new pools launched that kind of can service this new market that again is a massive portion of uni swaps Revenue so I kind of personally expect to see curve eat into a massive portion of that volume again this is kind of you know under on

the back of the assumption that the upgrade is successfully successfully reduces the gas costs used to run these trades but you know the curve devs are pretty based so I really wouldn't be expected to see anything else thank you yeah and if anyone else's eyes or ears kind of glossed over during all that stuff but it sounded really bullish just go follow when llama on Twitter and it's kind of like a high beta curve play I don't really know what the deal would deal with them is to be completely honest but I really like llamas and uh they're minting every hour and a half I believe until Sunday of next week so you got to get kind of lucky I think it's an auction style mechanism so you just gotta check back every hour and a half

and maybe get a cheap llama but it's directly related to the curved Dao Dan you probably know more than I do but the floor is around three eth right now so if you can snipe a cheap one it's pretty good R if none of that made sense too you could also subscribe to real quick research and we have a great report written by Dan about all of those fancy mechanisms that he just described and he really writes it down super well and even I could understand it which means you should probably subscribe but either way for dexas I think it makes sense right like I think everyone knows that that next fees are based at the bottom right it happened in traffic with stock brokerages it'll probably happen in crypto too because no one really wants

to pay 30 bips even on a long tail asset and so the fact that curves is an interesting mechanism to compensate LPS even without the presence of trading fees just seems really interesting compared to all of the other decks sort of designs out there well you guys are really making me feel like I need to tldr that so a quick recap stable coin and lending protocol gets you vertical integration and revenue diversification also with the ability to kind of compete in that Fearless race to the bottom that red was just talking about and the upgraded try crypto contracts allow you to permissionlessly pool any three assets of which ethusdc will likely be common pairings and that will help curve

compete greatly with UD swap in terms of volume dominance awesome well I think that's a good spot to call it over to our interview with Anatoly awesome everyone while we are joined by Anatoly the co-founder of Solana truly a man that needs no introduction but you know thanks a lot for joining us today Anatolia we're excited to kind of dive into a lot of the recent developments um of what's been going on with with Solana as a whole uh but also some of the more technical stuff as well as you know that's kind of the the sector that we'd like to dive into the most on this podcast but again there's been a lot of exciting news over the past couple uh you know days and even weeks so love to dive and dive in there first we've really seen this uh shift to a mobile

first perspective in crypto and you guys are really making an effort to capture on that with the release of The Saga the new phone powered by Solana and so I'd love to get your take on kind of the core strategy behind launching that phone um and you know this comes on the back of news of like uni swap leasing a mobile first app as well so we're really seeing that shift and yeah again just want to get your take on really the core strategy behind Solana launching this phone as you guys know it's really hard for web3 apps to actually get approved in the iOS and Google Play stores and oftentimes have to cut features and kind of hide what they're doing and it's the experience on mobile has been pretty bad uh like I would say like the transitions are kind of weird you

there's like every application you connect to there's like 50 wallets to choose from to connect to and then you're kind of dealing with this like weird context switching or having to copy a link to the thing that you want to do inside the embedded browser in these wallets so it's like it doesn't feel like the future of Finance is like at all you know so uh I think that's kind of like was one of the main reasons but uh you know one of the other ones was that simply I worked at Qualcomm for over a decade the stuff that we worked on there even you know when I was there we were already building a trusted display you can't trust a touch like secure display

and secure touch and the trust Zone that connects all three of those pieces to make sure that like when you're having when you have to build applications that have that like super high security requirements that Android can't even like look at the display or take a screenshot of it like all of these things the technology was already there that was like I think close to a decade ago so I felt that it's just really a shame that it's not by default in every phone that every device right now we just isn't like that perfect Hardware wallet that you want that's programmable with like a rich set of applications I can interact with it and stuff like that so that was that was the other reason I really wanted to see if we can like kind

of set an example and move the industry forward a bit but the main one is that devs really need like I think a third party store that's crypto friendly now something I'm really excited about is a backpack and xnfts with mad Lads minting uh tonight we're actually recording on Thursday April 20th for those uh listening on Wednesday uh six days later but is that gonna be integrated directly in Saga and are you pretty excited about that like can you kind of explain what x nfts are and why they're unique yeah xnfts are a way to deploy applications so it's kind of very similar to what an app store does what's really cool about it is that this is something that a standard that Armani built who's the founder of backpack

to for devs to build like applications that get embedded inside the wallet and you kind of need that because the wallet is a very kind of visceral experience of ownership and crypto like a what makes an nft is the fact that like Phantom and soul flare all display the same image for the same address and the NFD that's really what makes it if there's like another data structure that has the same kind of mathematical representation in the data there's like public key in the points of that image it's not an nft because they don't see it in my wallet this is like a very important like concept I think for adoption of like digital assets and the fact that you can

build a protocol with your own implementation of how to display and interact with that digital asset and then propagate it you know now to all the you know all the users of backpack I think is really really cool it's really going to I think accelerate the development of how users interact with web3 with new digital assets and stuff like that interesting and we've kind of been on a modular tear on this podcast like the last three or four weeks and you know in that conversation you often hear Light clients and I think it was Nick white from Celestia was talking yeah like I'm super excited to see how soccer goes it'd be awesome if you know we could get kind of like a light Client app store type deal on the Saga phone is this is the phone going to be very very Solana

focused in particular are you open to you know a lot of different other I'm a open source Maxi so like we just had to ship so like we cut out all features that were not that were like necessary and now that we have some Breathing Room uh are adding support for other curves uh so like I think honestly I like anyone that's using that to 519 curve should be able to use like add a wallet to it which I think a bunch of blockchains are anyways already so I think we want to add support for basically all the or whatever's missing We'll add support for it I think it'd be really cool to see like the entire industry start kind of like collaborating on a

of what these data structures look like for sharing applications on on chain but also moderation and all these other things because those are very much part of like the mobile experience we really don't want to be in control of any of it I kind of feel like users should be able to pick whatever you know application top 10 list they want right which is like where the moderation comes in and stuff like that hey everyone big announcement from the blockworks podcast Network we're launching a new show called Lightspeed and hiring Two Hosts to come build it with us lightsweet is a show for Builders focused on the use cases that will onboard the next generation of crypto users by taking learnings and inspiration from the garage days of Silicon Valley we really want to capture the perspective from

Builders because that's what the ethos of crypto is content experimentation and Relentless Innovation to build products that users can't resist if hosting a show like this sounds exciting to you then head over to the careers page on blockworks.com which we have Linked In the show notes you can also reach out to mirror Sam on Twitter to talk more about the opportunity but overall we're stoked about Lightspeed so if you think you'd be a great host please do not hesitate to reach out one other exciting development over the last week I think it was just yesterday actually helium making the official migration over to Solana can you kind of speak to that and uh you know where you're excited it is so cool I don't understand why it's not like the biggest thing in Tech because like 5G is uh is a really cool technology but it's very expensive to

deploy so like what most people don't realize is that 4G cellular networks those Towers have like a range of 30 to 40 kilometers but 5G the ultra wide band sells like only have a range of about 400 meters so it's very very short and they can't go through walls they can't go through trees so like if you were to just blanket United States Coast to Coast with 5G like Ultra wide map like cellular hot spots it would take like I think some people estimated like 10 million you probably don't need that many to cover all the users like because most users are concentrated and there's Wi-Fi in a lot of places but it's still going to take a lot of work to get you that

like I'm driving down the highway at 50 miles an hour like you know 60 miles an hour and I'm watching a movie and like my friends are all FaceTiming right like with somebody on the other side of the world also driving like 60 miles an hour right like it's just gonna take a lot of Hardware to do that and what's really cool about helium is that they've proven out that users can deploy a lot of Hardware like all over the world and do all that coordinating using you know cryptography in a blockchain so they did that with their like low-rand Network which is like built for iot so you can build like you know like iot devices that haven't seen battery life but now they're trying to do this for 5G

and they're competing with like the biggest corporations in the world with ATT and Verizon and stuff like that and I think they have a shot of actually doing it faster than them and if that happens I think that's like real proof that decentralization is like a net good for the economy it like touches like the very fundamental like thing that we like very fundamental piece of like the US economy could be built by decentralized networks that could be like I think groundbreaking for the entire industry so I'm really excited for them just to give you like a sense of like where things are at I think ATT and Verizon have about 50 000 uh like cell towers for like 5G I think helium is at

about 8 000. so it's not like they're a billion to like a thousand it's it's like actually like they have a reasonable shop it just started it's actually passing them and that I think in itself would be I think massive news yeah that would be a big win for honestly the industry as a whole uh but I'm curious so you know Solana's original messages was really about trying to bring like be the NASDAQ on the blockchain um but giving this more it feels like there's been some sort of a shift right and away from like that order book style like idea um into this more like mobile first idea so I'm just curious like how do you view the vision of Solana today well like the the focus on Speed and Performance um that is not actually going away we've

like tripled down on the investment there with fire dancer it's just when you improve cost and performance and like Fidelity and all these things that you need for like NASDAQ level Finance it makes everything else like just great it makes it easy but user experiences are like fantastic and you have like low fees and fast finality that's what users want and those additional use cases are not even like don't even come close to the amount of traffic that's generated by like oracles and Order books like pith and like open book and all these like like hero now like all an ellipses all of these like order book style dexes generate massive massive more number of transactions than the

users so my feeling was always like if we could get it to work and that like NASDAQ level it like it it works for Nasdaq and it's a decentralized Global permissionless Network then all the web 3 stuff is just gonna come for free I love that framework and thought process and one of the things you mentioned uh was fire dancer so you know having that unlock of diversification diversificating uh diversifying the network clients with fire dancer is a huge unlock and I believe that makes Solana the only other network uh with the second client other than ethereum so I'd love for you to dive in on you know really what fire dancer is and how it improves performance so this is a team out of jump training uh it was actually their Like Chief scientist is now

leading this effort it's a second implementation of the Solana client written entirely and see hyper optimized it's built on a technology platform that jump uses internally for data processing so they like handle 400 gigabits worth of throughput on the Slick software platform it's really really cool this is like what I dreamed of Solana should be designed like when I started but we just didn't have the resources to to like kind of build it from scratch that way but because they now see the end State they know exactly what we built on what the design is they can actually then like do these all of these hyper optimizations uh and the kind of like results are demonstrating are ridiculous

so like during break point the show they can do uh this is like commodity Hardware off the shelf Intel boxes it was handling 600 000 transactions per second throughput through like the transaction like scheduling stage which is really really cool Richard recently like tweeted their like demos of quick which is like the Internet Protocol that we're using for coordination they're showing like on four cores it's doing over 22 gigabits worth of like throughput so so like they're really really like uh like people like that are super into like like low level networking understandable like xdb XD XDP is and like

uh how to actually like take advantage of like all of this like low level Network Hardware should go look at the repo for finance or it's awesome yeah that thing just sounds like it's absolutely screaming and so that I guess that begs the question though of if it if fire dancer is so much more performant than the current Solana client like why does that like the you know you'd imagine every uh every validator is going to kind of switch over to running fire dancers so that kind of like remove the benefits of the diversification in this case so the key part for safety is that the runtime logic like the thing that actually plays the the transactions forward and The Logical State transitions but there's multiple versions of that so you could

have fire dancer be the primary client and then all the work of actually getting the block staying on tip and like dealing with shreds and all this other networking stuff it's done by fire dancer and then the salon is replay like client can basically work behind it and I already built Ledger and verify that the the same result is built by both clients and that gives you all the safety that you want because now you have two clients all the top validators that are like in the top 33 percent can easily afford running two servers that to them Hardware costs are pretty relevant so that's kind of the expectation there is that they will run one primary one secondary for a lot of the other folks that can pick one or the other so we'll

have like enough client diversity to give us those properties for liveness is a bit more complicated so with liveness you could have a system where like there's a bug something is getting saturated I'm like you know and running out of memory and you can switch to the other client but if you're switching from fire dancer to Solana you're probably gonna see a performance aggregation and maybe that's okay in those kinds of those kinds of events like it's better to lose some performance than to have like a liveness failure great appreciate that background um in terms of local fee markets I've seen you kind of tweet about that quite a bit and like isolated few markets can you just explain what these things are

how they work why they're important for the user experience yeah this has been like um this is what plagued us last year and we did have a chip like an option but to innovate which is really cool so the problem that you have with like blockchains that have low fees is that uh Bots that want to take advantage of an Economic Opportunity like an entertainment they will submit infinite number of transactions because very cheap to generate and send them very few of them will actually get through but they will basically prevent anyone else from being able to propose a transaction by saturating the mempool all the cues leading up to block production with dummy data so

what what happens is that in that environment you have like these malicious actors that are like economically incentivized just submit at one point we saw like literally like I think close to 100 gigs gigabit worth of traffic hit one of these nodes so these are like saturating the lion like the network cards at the data centers to handle the traffic going in and the majority of this fast fast vast majority of this traffic is dummy replicated transactions they even like modify them a little bit to like to break the deduplication like the simple deduplication filters so like all of the stuff is saturating the network and the validator can't drop them fast enough to get to the valid transactions

so you needed like a couple pieces fixed there one is you needed uh the ability to limit the amount of traffic that anyone can send and that's quick so we change from UDP to quick so now Bots can't just send 100 gigabits worth of data that gets sent some limited amount of course at that stage Bots will then like create a thousand different Bots that are all sending a little bit of data so to deal with that you have to actually limit the amount of connections that anyone gets based on stake weight and that kind of distributes that problem across the entire network so bot can't connect a million like civil notes to the leader that could connect a few and then maybe they can connect to other

nodes in the system but because we have a very large Network like over close to 2 000 stake nodes it becomes a much harder problem to saturate the entire thing so any anyone that has a connection to a stick node is guaranteed some some ability to send traffic through and because now you have like some guarantee that valid users get their traffic through to validate it when they receive this data it's not bounded that can actually sort it and drop it and this is where the localized fee markets come in so the way Solana works is that it's apparel runtime it can execute a bunch of transactions but if all the transactions are all hitting the same nftment or the same like liquidation for a borrow and a borrow and like soland or like a really

hot like I don't know Market in orca you can only fit so many at a time in a block even if you have other block space it's simply because these are all dependent on the same state like the same chunk of memory so those transactions are all have to be sequential there's no way to parallelize them because they're all touching the exact same spot in the state so you can kind of think it as like a hot spot like everybody wants the statement of demon that one chunk of memory that's identified with that nft man that stores like the state of that nft is the hot spot so you can only fit a certain amount of those so the way that the Sorting works is that takes the highest transactions by fee and fills them out up until that little State like that's a hot spot

fills up and the amount of work it can do but that's not the entire block so that's the amount of work that any single thread can do so like any single thread on your CPU or whatever it can take 400 milliseconds to work in a hot spot but then you fill up all the other spots after that so what that means is that like you can have a very hot nft mint like you're gonna see with like mad labs and the fees to access that might may Spike but that's not going to impact any other Marketplace any other transfer because they're simply not interacting with that state and that means they can be scheduled so because validators now have like a bounded amount of data that they can actually do the sorting and filtering the contextually that bucket all of these transactions by hotspots

and then take the top hot spots and put them into a block so it's like a you know like a Pachinko machine it's just like filtering all of these things into slots and then like taking the top ones and that happens to be your block this is a really exciting solution here and I'm curious if you can go a little deeper on how they identify hot spots and so like I guess in the example you gave right like mad Lads will likely become a hot a hot spot but that doesn't necessarily mean every other NF team that's simultaneously going on will be a hot spot as well so how does it like I guess how do you actually distinguish and group these transactions into these these like clusters if you will yeah so it's a pretty dumb greedy algorithm so it actually sorts everything by fee

and then going from the highest fee it starts tracking which accounts those transactions access so one thing that's very different about Solana than ethereum is that if you heard of axis list and ethereum it's that they're required on Solana and they have to be complete so transaction cannot access any data outside of the access list and a transaction has to have its access list declared so that means that looking at every transaction you could look at which access lists are being accessed by each transaction right so when you're doing this greedy algorithm you take the highest ones and then you start tracking which access list fills up first and as soon as that one fills up you don't allow anyone else to access it okay that

makes sense I guess I think I just we're missing a small piece because you mentioned how you know Salon is kind of designed with Hardware in mind and parallel processing is also another thing you've already mentioned so do you mind kind of explaining at a higher level uh how parallel processing actually works how it scales execution and you know the benefits or trade-offs associated with it yeah so the the tricky part is that the system like has to be parallel in every step of the way if you have any single Cube that's single threaded I think it's saturated you kind of lose benefit in the entire system so there's a lot of detail to explain there but fundamentally the key part that we did is that we parallelize the environment the execution environment the virtual

machine where transactions are executed so unlike evm you can think of every transaction in evm has like a global right lock on the entire state it can touch any part of the state it can do whatever it wants and then it executes and then the next one gets to do the same thing and it's all done very sequentially on Solana every transaction specifies exactly what stadium it's going to read and what state it's going to write so reads can overlap but rights cannot and this is where the the scheduling algorithm and how we do replay is that you look at all the sequential operations for any given State and you can put them all in one thread and then you take all the other non-level epic stuff and put them all on the the next

thread and so forth so then you have multiple threads executing all of the state transitions in the in the runtime so that gives it like that fundamental level of parallelism in the core execution part and then from there you have to make sure that the scheduler itself could handle the parallelism about the map Pool can handle the parallelism and like so forth because if one of those ever like fails you end up having like a sequential system that's bottlenecked so like one thing that folks don't realize is that this is the best you can do like you can't actually improve upon this it doesn't matter if you have zero knowledge VMS or layer twos or layer threes or whatever like what happens at databases

and in systems like this is that it's called the database hotspot problem everybody wants to access the exact same state so when you have like everybody in the world that wants to you know like flip the same light switch you know if you have 100 people that all want to flip the same light switch doesn't matter how how how many flight switches you can flip in parallel because right because everybody wants to flip only that one light switch doesn't matter if you like make the light switch flipping a thousand times faster somebody wants to go first and that causes a hot spot that means that you have to prioritize who gets to go first who gets to go second and this is where that localized fee markets resolve that is that people can bid to to access first now

the problem on ethereum and evm and mempools the way they work today is that if you have once one switch where 100 people want to flip it and then you have a totally unrelated thing like a Oracle update or or like a you know an nft mint they see the block getting filled up with high paid transactions that all want to touch that light switch they all want to go bid that one nft but Oracle updates still need to happen and people that want to liquidate still like want to catch those liquidations so don't they now try to outbid that entire use case to be included in the block and that causes this gas war between unrelated use cases because there's no parallelism in the mempool because there's no

parallelism in a runtime and as soon as you add that like you can't you're not going to fix that lack of parallelism with l2s because the same thing can happen in optimism you have an nft mint and optimism and a liquidation happening at the same time in the mempool there you're going to have a gas for and fees are going to go up so the only way you can do this is like actually with transaction isolation what Solana does and once you do that you have that parallelism established your only other kind of thing that can impact fees is just throughput and there you know add more Hardware every two years every 18 months roughly Hardware gets cheaper so things will always like an attempt to like get towards zero for the base fee

so it sounds like you're not exactly bullish on the idea of scaling ethereum through roll ups but I think a a common rebuttal to that would be like well you're gonna have app specific Roll-Ups you're not going to have an nft mint on a you know an app specific roll-up that's just a Dax or something like that how would you respond to that yeah so like Cosmos and like very application specific environments that only do one thing that's like a way to do that but that's not a like a generalized solution like that's effectively kind of very crude level of isolation you can literally do the same thing on Solana you can have like a program specific circuit that's like a ZK prover that's optimized for that one thing am I going to call it a layer two if I could

like it's not right like I think the the program specific stuff is cool and I think will work everywhere and that actually does solve a problem but it's not uh I would say like a general solution that you could say like this is the way to scale like these systems to like you know the entire world yeah I think that's a fair take um you also mentioned that you know you kind of see these gas Wars that happen on ethereum but ultimately that leads to you know ethereum's core value accrual in terms of like burn and you know Mev that accrues the validators and ultimately stakers um how would you say Solana thinks about this problem because that's how ethereum's tackling is tackling security I've heard you mentioned that you don't

think Soul the token is money and you don't want it to ever be money so I'm curious how you think security Works long term in first Alana I think those are totally unrelated Concepts I think security comes from operator security like it hacks it's the physical security of the network is the security of the network the token itself is virtual it doesn't actually do anything so it doesn't matter like how much of it is burned because an attacker can acquire the keys to the voting infrastructure for the network for free through remote code execution through all sorts of other attack vectors so the actual token itself is meaningless what matters is how many independent parties you have that have secure operations that maintain their state

of these systems and what's interesting you can have an unstaked node can provide tremendous amount of security like Circle they have billion like dollars of of on Solana they need to secure their node to make sure that that node is not corrupted they don't give a about how much stake is on that note it's totally relevant they have a business reason to make sure that their node is like secure and that is actually more value than all the stake in the world so fundamental security the system is that like the number of operators that are independent that give a that they have uh High secure like infrastructure making sure that that state is never corrupted so I think security is like a totally

separate subject like the gas Wars that occur on ethereum are extracting value from the use cases that are running on it from the users from the Liquidators from the actual math Bots if you think MAV is a net positive the gas is actually taking away from it and it is giving it back to the to the ethereum holders but I think over the long term there's no way that level of extraction survives Mass adoption it's just not how technology works like I think if you have like 100 million users they're going to be very price sensitive to fees so there's a few Roll-Ups uh two that come to mind eclipse and Nitro that are looking to deploy the svm as as l2s how would you

say your take is there considering you're generally bearish uh the idea of scaling a base layer with l2s I think it's awesome honestly like I think people are going to be building l2s they should be using svm like the runtime is better it's actually paralyzed like it I think offers like a better programming environment because it's using rust a modern language not only that it's using llvm so all the work we're doing to add like move and solidity support is available to all those folks so like myself as like an open source Maxi nerd it's awesome like it's licenses under Apache license for that reason like we want people to go take the code and go build products with it now whether I think that that environment is going to

be as good as Solana later one that's a different thing but all the work that they're putting in ultimately feeds back to you know Upstream so like if they find bugs if they find improvements or or like build better tooling I think all of that part all of that is like I think I'm not good for the entire ecosystem in terms of security as well back to that topic I think a lot of ethereum uh people would also say well you don't really have any at-home validators when it comes to the Solana it's mainly people you know with large server access or whatever it is and the resources to actually spend the hardware uh spend on the hardware so how would you respond to that in in the case of like a regulatory backdrop or or

Crackdown I should say or basically any scenario where the network needs to fall back to at home validators um so like one thing is that like people way overestimate how hard it is for the government to crack down on at home validators like first of all the ice bees that you're connected to Comcast ATT for home connections those are the all those companies are effectively extension of the federal government they're federally subsidized they have to get their licenses from the federal government and the amount of users that they would have to shut down is very very small compared to their total user basis like compcast is not going to care if they

kill every home validator like so you're actually like like if there was like a real regular sort of Crackdown home validators in in the US will not survive and very unlikely I would say a majority of the world um so like from that perspective I don't think there's any advantage anecdotally there's already home validators on Solana simply because you're starting to see like one to ten gigabit like network connectivity at home just just for the rollout like just from the general upgrades so I think that that's like kind of happening on its own but I personally believe that for these systems to scale they have to be highly available systems and when you look at Solana it's an over 200 individual data centers like the network

is in more data centers than most networks of nodes so it's very very much distributed around the world um I personally believe that like crypto if it's gonna succeed it's gonna succeed in the United States and I don't foresee like a future where like us has a massive Crackdown of like there's even a point to a high performance blockchain like it's basically gonna be like Bitcoin like uh monetary maximalist like waiting for the the Rapture of the US dollar collapsing yeah Fair points there and so you know as Sam mentioned like there's definitely that you know Solana like catches heat for that and the way I'm like starting and also I say I

mentioned like we've been on this modular tear for a while and the ideas of like a shared sequencer system but you know in my mind when you start adding world after rope after roll up onto the same shared sequencer the hardware requirements to run and participate uh in this in that sequencer shared sequencing system is eventually going to hit a point where you know is it going to be drastically different than the requirements to run a salon a validator so I'm curious like what your thoughts look around uh like the shared sequencing landscape yeah these are like even so when people say it inherits the security of the layer one they don't actually speed that like me is a layer one user that I magically have layer one security I'm going to roll up

like the roll-up could could run invalid State Transitions and if there's no full nodes there's nothing me as a lawyer one user will be able to do so me as a layer one user to get pure trustless execution you have to run my L1 node and my L2 node to make sure that if L2 causes a fault that it's submitted now it's a little different with zks but still the provers are like inscrutable you have to run like a classical version of the same computation to make sure that the prover did screw up like there's trusted setup there's all these other issues that may be solved in the next 10 years but not not right now so like if you consider that that you're running multiple nodes they have to be high

availability systems you could run a Solana validator right now for 350 bucks a month it's like you're looking at the cost effectiveness of each now like I think things are much more interesting with like like clients and things like that which systems are more more suitable for like client proofs and like getting the idea that when you have at least a large user base of like node providers that are running these for business reasons that at least one of them is honest and that is enough to guarantee like shared security of the system this is what you get as a layer one user this is what you should be getting as a layer 2 user as well without actually running any notes as well but you need like either you can do this socially you can

literally go into the channels where people that run these systems like work and like keep an eye on things and like you receive a transaction and you wait like eight hours nothing blew up right like nobody's claiming that there's a fault it's probably okay but it's better when these systems are automated with protocols and you can get like these guarantees within like 30 seconds or a minute um and I think from that perspective when you have like a system like Solana with a bunch of big block with a bunch of big validators that are trying to maximize that number of full nodes so at least some some set of them are honest working with like clients that provides a very high like like very high security for the vast majority of users and that

is I think very competitive to low cost validators but a bunch of effectively High compute cost L2 nodes but users still need to have protocols to get these like shared security guarantees the trade-offs are like from my perspective it's actually easier to analyze Solana because you can look at it and say okay there's like 2 000 full nodes a bunch of them are businesses that are running this to make to like that have valid reasons around this not just like getting tokens from staking rewards like one of them is going to be honest right so and all I need is a protocol to make sure that uh like that signal gets propagated if something happens

so uh long term I think like the like like light clients will kind of eliminate the Gap in what the normal user gets as security between all of these networks and that performance and like finality will be like the main things so I'm curious how do you think about transaction fees as a whole right so if you look at you know Solana it's known for its high throughput low uh low transaction fee model and like to the end user do you really care if like once a once a transaction is say below one cent like is there any difference if it's like a hundred of a thousandth of a cent at that point are you like over optimizing for you know like Mev at that point well like the central limit order books

the market makers and the pith oracles they care about the fees because they do submit like over 20 million transactions per day so that starts to add up I think they would like to see like you know the financer team like their goal is to get us to like that like over a hundred thousand TPS like steady state that means that fees can even drop even lower and they can put more data on chain for people to trade out and to that that's like that would be really cool because that impacts their normal business like if if there is like a public data store of all the NASDAQ trades that was that's not a big pipe to Solana that would be like I think that would impact their bottom line just and a lot

of like trading firms bottom line so I think I think reality is that like as the system scale machines are going to be accounting for 99 of the volume users are going to be like users are going to have very low fees like the cool thing is that like a user can pay 10 times more to outbid everybody like all the Bots and get very like high level of service on Solana and it's still a like less than a fraction of a sun so like you can 10x your fees like always like as a as a salami user and you get highest priority above like the average transaction and it'll have no impact that that I think is that that kind of level of performance is what we really want to provide is that like users have the option to get really fast

service still at a very very low cost but the network is a data Rich environment with lots of participants like you know doing stuff right no that makes a lot of sense as well and I want to kind of take the conversation to dive in on on the state of Mev on Solana so I'm really curious just to kind of get your your overarching thoughts before we dive into more of the specifics like how what is the the landscape of Mev on Solana today um so Judo I would say is like the leader there so they have a stake pool I think it's over 20 percent of nodes are running a Cheeto client so they have like a Searcher Network that's optimized for you know finding those like opportunities that's and then submit bundles into the network uh it's awesome because they're like creating client

diversity they they have their own build of the salon of validator so now there's not like a single failure point of where you're going to get your source code from which is I think really important and they're very smart engineers and they're very much performance sensitive so like as many of those folks as we could get looking at the code and like optimizing and testing I think ultimately results in better outcomes for users and uh it's so it's like having that as a transparent part of the ecosystem I think is very important because it I don't think there's any way to like really get rid of mpb uh so making sure it's transparent competitive is what's going to result I think in the best experience for the users so do you think something like

enshrining PBS proposal Builder separation is like a a path that you'd like to see Solana pursue yeah we've had this idea uh we called it multiple leaders per slot like it's been kind of part of the Genesis of the Solana design is like that if if we have multiple block proposers proposing at the same time then the user can basically send their transaction to the one that's offering the highest math rebate so these proposers are then like competing right for user transactions and they're like an average seeing X returns and they're like hey suddenly all your flow will give you a rebate like that should be like the end state of Mev is that the proposers are very much optimized like I don't know maybe

around a massive GPU forms trying to figure out what is the triangulation like bundle like best best outcome so like that's something that's been I think part of our intent like from from day one there's a proposal right now like uh we call them cmds on the in the Solana Foundation repo to uh that actually is pretty close to the proposer build a separation that you see on ethereum now I saw you put out a tweet I can't remember when it was but you were saying that daps could actually recapture some of this Mev is that exactly what you were just referring to or that looks somewhat different so that's a bit different so like one of the things that we see is that uh there's a lot of uh opportunistic

arbitrar spam in a network so like uh basically whatever if you have like Financial State in these systems uh and you have X amount of capacity right or x amount of like access to that state Arbitrage will fill it like the entire capacity of that state if the transaction fees are low enough and that's actually fine like we don't really care so they're like submitting a bunch of transactions 99 like percent of those ARB transactions fail because they're opportunistic hey if this price is X just happens to be X and I get lucky then I know this is a profitable transaction so why don't I submit it because the cost to do that is pretty pretty is low um

one of the ways that we thought that that could benefit applications is if they could charge these transactions just for taking a right lock to the state so like if a transaction wants to access mango that's an orb that's like hey I'm trying to see if there's like a price on Mango that's that I want to take but if not the transaction fails that just just even asking that like querying that state and taking the right lock it has to pay mango a certain amount then Bingo can tune their fees to basically like capture this ARB like this opportunistic ARB returns mango itself can capture it and give it as a rebate to users that are creating valid

like uh bids and asks that are adding liquidity to the system so from a very high level you can think of it as like your traditional uh ESN like exchange that gives people that are creating liquidity a rebate and those that are taking liquidity paying a fee so this is kind of like the the system that we want to build in like create at the low level because that's very hard to do at the application Level itself because from a like a high level like outside the app I can submit my ARB if the application doesn't do what I want I abort the transaction it means that the entire State transition is canceled and bingo doesn't get anything so this is kind of like what we want to

undercut is the ability to do that uh uh by by these like opportunistic arbs and then they can actually do this like Balancing Act okay we can asset our feasel higher and that means that there's going to be fewer arbs or at least the arbs are picked more and then we give a rebate to everybody that's adding liquidity and that that's I think creates like a better environment for for users and Traders I love how you're approaching Mev in this fashion right like instead of worrying about the money-ness of Soul yeah the Mev is flowing back to the application and the users that are generating it um is that is that is that like broadly your thesis there is instead of the token rather focus on the the users and the creators of this Mev yeah I think ultimately like what's good

for the creators and the programs and the users is good for the network I think that's actually I think better uh more sustainable approach for growth because like while it seems like you're getting a return right like oh more eth is burning as soon as everybody stops using it doesn't matter how much ethi burn prior to that like it's irrelevant like it all like all that burn is like useless it's not like a battery you're charging uh does that make sense like as soon as like people no longer use it the network is dead that that's actually what happens right like so what matters is like how do we create an environment where like these businesses can grow grow as fast as possible attract us the

most users they can and from that perspective I feel like the protocol should be like the least extractive like anything in code that we can do like to make it more efficient to give it more capacity to like avoid like uh unnecessary bottlenecks right like I think that all of that is on the table so more broadly speaking I mean I think you know there's a good chance that the future is is multi-chained so like I guess what is your guys's strategy in terms of communicating with other virtual machines and in a secure manner I think the only chance at surviving a multi-chained world are the ones that believe that they're the single chain [Laughter]

to survive you have to go for for gold for Glory right I love that uh the way that we're approaching the multi-chain like multi-vamp thing is that uh solanus LV like using the lvm stack so the byte code that it's using is a backend to lvm so like there's a project to uh compile move through that stack and what's cool about that is it's forcing us to add like next Generation features into the back end uh where like now the bytecode is typed and you can do type inference between different programs and then combine them during the like the jit stage and then you can have uh actually like calls between

these programs that don't need to like have any safety checks that you have a guarantee based on the byte code that like when I call the token program that I'm passing the right data and that means like you eliminated a whole bunch of like runtime checks and like things get faster so that's been really cool like just having move as a forcing function for us to like add all these features that like the kernel the Linux kernel folks added to to Berkeley packet filter a couple years ago um so that like that's basically our approach lvm is very general purpose compiler uh I can't imagine it like not being able to do something like that that people come up with in the virtual machine space there's a solidity project

too called solang that's like I would say fairly mature but still takes a lot of like uh like unexperienced engineer would be uh would need to be able to use it because a lot of the kind of like the basic tooling like hard hat is still missing but if you're like an experienced solidity dive you could get going with it I honestly love that mentality that that that's that's a good this is the right framework to be like attacking this this idea but I'm curious what uh like what do you think the Solana ecosystem is like missing today as far as applications go what would you love to see get built there I think like what it's kind of what all of crypto's missing is there's no like application that we can point to and say this is general consumers using crypto

and getting value out of it and these are the business models that are creating value for the world it's just not there's nothing here so like I'm like we're trying like to attract folks that are have that mindset that really want to unlock like value for the general consumer helium I mentioned is one high mappers another one uh folks don't know what that is they're basically competing with Google Street View they have Hardware cameras that are on devices people are effectively mining like tokens by creating these constant of constantly updated streets and it's a really cool product because they're uh their updates are coming in 200 times faster than Google's just kind of crazy like because they

have like these drivers that are constantly Mining and uh they're 60 million kilometers of Road in the world they're over a million already so they're moving really quickly right like like two percent of the way there for the entire world I think that could be like very like substantive like technology built using decentralization that we can point to and say hey look this is competitive with the you know two trillion dollar company but it's still not like consumer stuff it's like infra similar to helium wow like my like what I'm excited about right now is like um so like backpack and dialect I think are new approaches to crypto where they're trying to

basically build something similar to WeChat where you have social and finance all mixed together and they're taking two very different angles on it I think dialect is very like mobile first chat app uh backpack is a wallet with a lot of like unique and cool like interactions you can do like including social uh but the goal there is like if you have like a moderately successful social network that's struggling to raise money in Silicon Valley for like 20 million users would be the biggest success in crypto we can all them point to it and look look there's real consumers they're doing stuff they're making payments they have like simple returns of unlike you know stable coins or something like that right like it works like the whole Loop

works and there's and it's growing like if we get one of those successes I think that would be revolutionary and like once you have like that many users that are actually using this stuff on a you know monthly daily basis then I think the business models that make sense will just be become very obvious and fast to optimize and you'll see that like next next Generation like all of that is missing from crypto like it sucks like it's been you know I've been working on this for over five years and we still can't point to like that one use case that's like that's the consumer use case that's the winner yeah Sean agree there we me and the the rest the research team have been preaching on that that front as well like you could maybe argue stable coins as a potential killer use case for crypto I think that's probably

the only one but at the application layer it's few and far between um in terms of backpack too I was very excited to see uh 2048 Flappy Bird like some of these classic web 2 games in there like kind of obfuscating the difference between you know web 2 web 3 I thought that was super cool and super unique but as someone who's so bullish on their own project which I absolutely love I would love to hear um the bear case like at what point would you admit defeat and be like damn like I was completely wrong on my thesis on how to scale blockchains oh man well I don't think I'm gonna be wrong on scaling part I think we have to for me to like question it I'd like to see ethereum and all the l2s combined to even match what

Solana does from applications today like reality is that like Solana does more transactions and all of their scaling combined right now uh I think for that to happen you need like thick prototank sharding might get close to that but uh I'm not 100 I'm doubtful because it's just one of the pieces you need to like really optimize the map pulls as well and I don't know how much work is being done there and that's going to start touching the execution environment and like unless you do all three of those pieces you're gonna have like these weird bottlenecks that they're not gonna quite work well together but at least that looks promising they're like focused on increasing the data availability layer

and that's ultimately the bottleneck is the bandwidth between notes can you can you maximize that um so like I think Proto deck sharding might might get at least closer um I guess I would be impressed if like the posting sharding if the throughput of that system was anywhere close to firedancer like I think I would be I'd be like okay this is like very much very much a valid scalic solution I think probably my guess is that fire dancer is going to be like 100x [Laughter] it's just like it's really really hard to compete with like a heart like a an

optimized server at a data center with like 10 like 10 to 100 gigabits worth of available bandwidth and like it's very hard to compete with that system right it has all the cores all the raw resources the only bottlenecks right now on Solana are like are stupid code that we wrote as fast as we could to launch like that's not fully optimized and the fire dancer guys don't have that like constraint they get to they're like actually building it the way it should have been you know from scratch like from the ground up so like like software is the only bottleneck right now the hardware is like ridiculously fast and not expensive we're not talking about like a hundred thousand dollar servers this is

literally like 350 bucks from latitude.sh is like you know multi like very large core system that can do 10 million of Ops per second on a single core like this is how fast Hardware has gotten today so I I like I have my doubts that like the even posting sharding that'll be anywhere close to like a well-optimized not work like Solana running on good Hardware well no doubt it'll be a fun a fun race to watch and uh I really appreciate you taking some time today and coming out and uh sharing your thoughts on really your your viewpoints from uh the boots on the ground Vantage we we would love to hear it from that perspective um so I'd love to give you a chance to kind of tell the audience where they can

learn more about you Salon as a whole and anything you want to kind of leave them with on uh what to what to be watching out for go buy a phone at salonabubble.com it's so easy to learn to advertise other than that go to salon.com you can learn all about Solana what we're doing but yeah if you want a a phone that is like actually opens up web3 to developers to so they can do whatever whatever they want and like build out some experiences and applications go check it out it's actually like the best Android build I've ever I've ever seen I like the titanium shell my guess is that like apple is gonna copy it for their next device yeah I as someone I've used the phone and I gotta say seeing the DAP store directly next to the Google Play

Store it was such a zero to one moment for me everything felt real again and it was I I'll definitely recommend go get this phone but again thank you thanks a lot of Anatoly appreciate it