


Making test-time compute go brrrr @frok_ai @LlamaRisk Stables & veTKN connoisseur @vyperlang enjoooyer @WenLlama #4


There are countless ways to address state growth if the first/second order effects would actually be understood Who is responsible for much of the ongoing state growth? Maybe it is PropAMMs, dozens of them doing hundreds of writes per block Solution - Global storage contract or opcodes, maybe a good place to introduce state expiry there. Make SSTORE expensive the first times, and cheaper as it warms and remains in use (SLOAD’ed). More state would end up shared too. What drives sstore cost? (at exec) Book keeping r/w and reverts. Or uncertainty of completion. Solution - Allow, bulk read / writes. And charge less for that. See what monad is doing Why do devs seem to “waste” storage? - (un)packing state grinds on code size / gas, so devs use more storage - devs with demanding use-cases fall back to diamond pattern galore. Effectively causing more state growth as more contracts are needed + cross contract call codec overhead Solution - 64kb max contract size. Charge more here, it’s ok. Jumpdest still fit into two bytes, absolute no brainer change 5x SSTORE is the laziest and least informed way to do it
These fuckers really don’t even use their own chain. 5x on SSTORE costs.

type shit
one of the most underrated asymmetries right now is that slop is totally fine for the backend and basically fatal in the frontend. if you’re backend is slop, the app still works. if your front end and overall experience is slop, the *product* does not.


okay public speaking is hard, apologies for the folks who showed up to watch me overheat and bomb, so here’s all the stuff i meant to explain right now you are being asked for a near 100% prepayment because neither you nor the cloud is considered credit worthy, and so nobody can get debt. As ai scales, we are asking the credit markets to take on a massive single sector deployment of capital. More money deployed, more scared everyone gets. this problem will flow down into the token layer. Every inference provider or model lab struggles to be credit worthy enough to get compute, and their lever to improve their situation is to lock in big prepayments from their customers or long term contracts from credit worthy customers. So this problem doesn’t just effect compute buyers, it effects token buyers. We’re entering a world where fortune 500s can very quickly deploy huge token spend ON CREDIT where the startup will need to raise capital. Startups got to exist because human capital could leave a firm and get ahead of the big firm by exploiting a combination of incentive misalignment, speed, and taste. You probably have the hidden assumption that the fortune 500 is slow and incapable of competing with your ultra fast startup. This belief does not necessarily hold in the case where the agent does the thing. the “market” is working on solving it. We know, because we are the market, and we’re working on it. sfcompute is working on ways of firming up the collateral of the loans, which makes it safer to lend to a startup. this is basically the whole reason we exist, the problem everyone is experiencing now is the problem junelark, our little ai lab years ago, hit however, we won’t succeed in time. so, in the short term, startups and venture firms should learn to protect themselves. we are mostly watching startups and investors make unforced error after unforced error, because this is not a simple subject. first, venture firms need to understand that they are in the compute business now and deploying capital to pay for a 100% prepayment for 10 startups where 6 of them die and 4 of them make 50% margins on that capital is basically locking in non-venture returns. many funds, via low understanding of the space, are wrecking their funds. there’s an economic solution for vc’s and it is owning the compute and trading equity for gpu/hrs. If the company dies, you keep the residual value. Each of you are looking to capture the next anthropic or openai, but the only way these companies scaled was by trading compute for equity. meanwhile, startups have three options in a credit crunch. The first is to raise a wild amount of capital, enough to pay for your deployments with far larger prepayment or to survive until things chill out. theres a reason we’re seeing a number of multi-billion dollar rounds simultaneously. The second is to design your business model to support variable priced tokens and compute. As the financial ecosystem builds around compute, you will be able to trade “lower up front payments” for “variably priced compute or tokens”. Your business will get tailwinds from the infra under you if you setup your business model such that YOUR customers have variable prices and you’re just adding a % take rate on top of the underlying floating price. Every year, you will become more and more capital efficient. The economic winners in ai & inference are firms that make lower and lower margin every year, but become more capital efficient & scale faster every year, making their gross profit increase. The third option is asynchronous or interrupted products. If you have a variable cogs, but a fixed rate card for your customers, it can’t have 100% availability or you’ll go bankrupt. But it is totally possible to design your product such that customers like this.
If you or your portfolio have not gotten up to speed on the current credit crunch, SFC is hosting an emergency Q&A session tonight, curtsy of our friends at planetscale. luma.com/x2m0vn3r

Even with missing rebates and outstanding propAMMs integrations you should see a 1 to 10bps gap at 10-200 eth. We've added 13 new @aeroxyz pools to the router today. Widening the bps lead, at the cost of a higher gas footprint.
try 10 to 200 eth. ---- btw: side-by-side by hand is really not the best way to compare 1. You get surplus in any tx and block going forward. usually is not passed on to the user; except over at @CoWSwap (and still not to the full extent). The most accurate bench is a realistic 4 to 14 sec delayed sim, mid block. That is where it shines most with wide gaps. 2. off-chain aggregation is always a delayed view: -> Sequencer -> Node / Indexer / ETL -> Receiving request from UI -> Solving then simulating -> Sending back to UI with an on-chain router you fire the simulation right away and get a fresher view / quote hence, it will always look outperforming when market goes up, and the opposite if the market is going down. that aside, if some propAMMs would be kind enough to match whitelist / fees aggregators get, for an arguably less toxic flow--amortized risk over all liquidity sources when routing BBO; that would help to close the gap (in a fair way)

try 10 to 200 eth. ---- btw: side-by-side by hand is really not the best way to compare 1. You get surplus in any tx and block going forward. usually is not passed on to the user; except over at @CoWSwap (and still not to the full extent). The most accurate bench is a realistic 4 to 14 sec delayed sim, mid block. That is where it shines most with wide gaps. 2. off-chain aggregation is always a delayed view: -> Sequencer -> Node / Indexer / ETL -> Receiving request from UI -> Solving then simulating -> Sending back to UI with an on-chain router you fire the simulation right away and get a fresher view / quote hence, it will always look outperforming when market goes up, and the opposite if the market is going down. that aside, if some propAMMs would be kind enough to match whitelist / fees aggregators get, for an arguably less toxic flow--amortized risk over all liquidity sources when routing BBO; that would help to close the gap (in a fair way)
If you're trading eth on @base you'll get better rates using @frok_ai. true till 150k but from what I've heard that should increase soon @0xngmi

What makes routes go stale—an interactive (and gentle) intro to how Frok’s JIT-router turns volatility risk into positive surplus. frok.com/blog/jit-router/

Tokenized equities are so fragmented by issuers and liquidity. One company (E.g., NVDA) can have 5+ wrappers , each with different underlying structure , trading hours and live on different chains. If I traded stock and wanted to buy a tokenized stock onchain, I would find it confusing too. We still have a lot of room to improve.

I don't have a horse in this race but there is exactly 0 (zero) point in comparing whatever the UI quotes and judging execution quality based on that without knowing the backend infra / state lag of the providers etc. Anyways... you guys should probably take a look basescan.org/address/0x00000…
Decided to test this out. Went to matcha, which defaults to USDC -> ETH Typed in a 1,000,000 USDC Uniswap: 417.5 ETH Matcha: 417.3 ETH am I doing something wrong?

Shock-and-awe
Today we unveil our first JIT router for ETH/USDC, available on swap.frok.com — setting a new bar for optimal aggregation — operating exclusively on-chain, at execution time. The pre-release router executes BBO against a ~300m TVL hub on @base , spanning over 100 liquidity sources including propAMMs. We went top-down with ETH/USDC as the first and most traded pair with the highest routing complexity. More pairs and chains to follow in due time. Integrations and feedback is welcome. We would love to hear your thoughts as we stabilize the interface. Bls dm. 1/4

broke: chasing ev in +1 woke: anticipating same block ev bespoke: anticipating +1 chasers, and doing the opposite, assuming its impact and going for same block ev in +1
lots of same-block DEX backruns on RH chain, anyone know if orderflow auctions or private flow deals going on?

Hard to fathom the climb it took to build this. Solving for same pair N-pool, then multi-path, then trying to win the latency race, eventually realizing off-chain is a dire path, with no end. “fuck it just do it all on-chain” then turning into a superoptimizer and a yul fork to get around compiler limits, enabling probably the most cursed golfs one can do in the EVM
The last on-chain router you'll ever need And the first on the EVM to execute at scale, solving splits entirely on-chain Delivering perfect execution across up to 120 liquidity sources—in any block, at any position Landing soon on @base, covering major pairs and stocks

The last on-chain router you'll ever need And the first on the EVM to execute at scale, solving splits entirely on-chain Delivering perfect execution across up to 120 liquidity sources—in any block, at any position Landing soon on @base, covering major pairs and stocks

argot was spun out of ethereum foundation with a mandate to maintain ethereum's core programming languages and developer tooling. then it immediately begins to launder research as if it was core infrastructure maintenance. if you read their blog, they spend a lot of resources on fe, a language that has been "emerging" for over 5 years. they have long plans for fe, while the language itself has seen zero adoption and zero production use. their long term goal is "non-trivial contracts in production-like setting". meanwhile vyper is actual production infrastructure. it secures real protocols, with real users and tvl, and real audit surface. curve, lido, yearn, frax, velodrome all use vyper. yet vyper lives grant-to-grant, while argot started with a $16.6m check, about as much as ethereum started with. argot doesn't disclose how much time and energy it spends on the fe fantasy versus solidity, sourcify, hevm, or other genuinely core tooling. but clearly this pet project abuses and stretches the mandate. even though it's a programming language, by no serious measure it's "core". it should spin out and try to survive and prove demand independently. production compiler maintenance should get baseline funding before speculative language incubation gets considererd. vyper is in good shape today despite the ecosystem, not because of it. and it still does not sit right with me that resources keep getting misallocated away from the compiler people actually use. ethereum keeps saying "public goods", then funds the toy compiler like infrastructure and makes the production compiler pass the hat. that is not stewardship.
