Imagine a DeFi lending market that lives on an Ethereum rollup but liquidates loans straight into mainnet liquidity.
Even better? Its markets could be cheap to launch and cheaper to use, with collateral sent its way landing almost immediately. Plus, the rollup underneath could prove itself to Ethereum every 12 seconds, even on its worst day.
The hypothetical I've sketched above isn't possible today, sure. Yet such a system is much closer to reality than it was a month ago, and the pieces for it are coming together outside of Ethereum's hard fork pipeline, i.e. the slow, deliberate process that changes the chain's core rules.
In fact, the main pieces necessary for this sort of build all took a step forward this week. Consider:
Solar, a new Solidity compiler from Paradigm
Rollup X, the first network built on the Ethereum Economic Zone (EEZ)
Fast Confirmation Rule (FCR), via a notable new beta
evm.fast, an open contest for proving Ethereum blocks
Solar, a new Solidity compiler from Paradigm
Rollup X, the first network built on the Ethereum Economic Zone (EEZ)
Fast Confirmation Rule (FCR), via a notable new beta
evm.fast, an open contest for proving Ethereum blocks
Let's break down the basics of each of these legos, then put them back together to give you a better sense of what I'm talking about.
First up, Solar is a new beast on the block. In short, it can make Ethereum contracts cheaper, and it's arrived without an EIP.
That's because on Nov. 2nd, Paradigm shipped v0.3 of Solar, its new Rust-built Solidity compiler (i.e. the translator that turns Solidity into the bytecode Ethereum runs). It's the first version that actually produces bytecode, and developers can try it in Foundry with forge build --use solar.
Per Paradigm, Solar-compiled contracts cost ~26% less to deploy and ~5% less to run than solc's, the long-standing default.
Those numbers check out in early DeFi tests, too. For instance, in its OUSD repo, Origin reported 35% cheaper deploys and ~4% gas savings per call with zero code changes. Not bad!
The catch for now is that Solar's own README says not to use it in production just yet. Savings aren't uniform, either. When Ekubo moved its development branch to Solar, 166 gas measurements fell but 12 rose. Either way, though, there's big potential here.
Secondly, there's Rollup X, which is pointing the way to seamless crosschain Ethereum UX.
If you missed my explainer back in March, the Ethereum Economic Zone (EEZ) is a Gnosis and ZisK-led effort, co-funded by the EF, to give Ethereum and rollups synchronous composability, i.e. one chain's contract calling another's and using the answer in the same transaction, with everything reverting if any piece fails.
At Devcon this week, the EEZ launched Rollup X, its first public test network, with demos to match. In one, CoW Swap filled a trade signed on Rollup X with liquidity sitting on mainnet, and the tokens landed back on the rollup in the same transaction. In another, Rollup X checked a post-quantum signature so a mainnet contract could act on it.
The tech is still maturing here, to be sure. Real-time ZK proofs are expected "over the course of 2027," and Gnosis says the network "is not yet meant to hold significant value." Regardless, Rollup X is an early taste of what's to come.
Next up, the Fast Confirmation Rule can facilitate deposits (and more) in seconds.
Today, most L2s and bridges wait for finality before crediting deposits from Ethereum, which takes ~13 minutes. Per the EF's explainer, FCR instead counts validators' attestations (i.e. their votes on blocks) and can treat a block as confirmed in about one slot, or ~13 seconds.
FCR isn't brand new, as Gnosis's bridges have used it for Ethereum-origin transfers since Oct. 8th. What's new is easier access. On Nov. 4th, QuickNode put FCR in beta behind the safe block tag apps already call, which should make it a lot simpler for other teams to follow. As Ethereum researcher Julian Ma put it, "FCR adoption is now unblocked."
The crux here is that FCR assumes at least 75% of stake is honest and online and that attestations arrive on time. If not, it falls back to finality, and in extreme cases a fast-confirmed block could be reorged.
Lastly, evm.fast is tackling proving's worst case.
The whole "prove it to Ethereum in real time" vision depends on provers keeping up with every block. They mostly do: per Gnosis, ZisK now proves 99.9% of Ethereum blocks in under 10 seconds. The trouble is the long tail. Nobody can yet prove the most work a single block could take, so a nasty enough block could, in theory, blow up proving times.
evm.fast, which went live Nov. 2nd from the EF, zkSecurity, and Verity Labs on the Yukon research platform, turns that problem into a leaderboard. Entrants compete to prove, in the Lean theorem prover, a hard upper bound on how many RISC-V steps (i.e. basic machine instructions) it takes to validate any Ethereum block.
The current record on this leaderboard is infinity, so any finite proven number would take first place at the moment.
These pillars are all powerful in their own right, but when you put them together you can unlock new kinds of apps onchain. For example, let's go back to that hypothetical next-gen lending market idea.
In this vision, the pieces I've been discussing could fit like so…
Solar for launching markets. Many lending protocols deploy fresh contracts for each new market, so cheaper deploys would compound as a protocol grows. And since interest-rate and liquidation math runs on every borrow, repay, and liquidation, even single-digit runtime savings add up for users.
EEZ for liquidations. Lending markets on newer chains are risky because there often isn't enough local liquidity to sell seized collateral into during a crash. With EEZ, though, a liquidator could sell collateral into a deep mainnet pool and repay the debt on a rollup in one transaction, with the whole thing reverting if any leg fails. The rollup could then more safely list collateral whose real liquidity lives on mainnet, and arbitrageurs could keep its prices in line without taking on bridge risk.
FCR for incoming collateral. Not every dollar will arrive through a synchronous call; plenty will come from exchanges and other chains. With FCR, the market could credit those deposits in ~13 seconds instead of ~13 minutes, so capital sits idle less and liquidity can rush in when rates spike. The DeFi-native twist? FCR-confirmed collateral could count at a discount, or under a cap, until it's finalized, so a rare reorg can't leave the market holding bad debt.
evm.fast-style bounds for bad days. Liquidations matter most when markets melt down, as that's when blocks are fullest and an attacker has the most reason to craft one that's slow to prove, stalling the rollup's link to mainnet right when liquidators need it. A proven worst case, like the one evm.fast is chasing for Ethereum blocks, would let the operator buy enough proving capacity to land a proof every slot regardless. "Usually fast enough" can become "provably always fast enough."
For users, the upshot of this kind of app would be simple: borrow against assets whose deep liquidity lives on mainnet, move collateral in within seconds, get tighter prices, and pay a bit less gas. And all this without a hard fork required!
That lending market idea is just the first hypothetical that came to mind. You could swap the lending market for something else, and the same pillars could be applied similarly.
Take a post-quantum treasury app for DAOs, for instance.
Signers could approve payouts with quantum-resistant Falcon keys, a rollup would handle the expensive signature checks, and the mainnet treasury would then act on the result in the same transaction, just like Rollup X's demo.
Additionally, mainnet would pay only a slice of the ~5.5M gas a direct Falcon check would cost (about 495k in the demo, per Gnosis). Solar could trim the verifier's costs, FCR could speed up incoming contributions, and a proven proving ceiling would keep the whole thing chugging along smoothly.
We can revamp exchanges, payments apps, onchain games, and beyond with these new tools. My grand point is that the menu of what's buildable on Ethereum just got longer, and without a single protocol change.
It's no secret that Ethereum's hard fork pipeline is a bottleneck, and that's fine. Making core changes to the chain should be approached slowly and methodically!
Yet it's also true that the Ethereum ecosystem can ship around this bottleneck to expand what's possible, and this week has shown us a bright glimpse of this DIY dynamic in action.
Nowhere will this trend matter more than quantum. Ethereum researcher Justin Drake has said the technology for a post-quantum Ethereum is no longer the main holdup but rather the All Core Devs (ACD) process.
If every client team made quantum the top priority, Drake suggested it could happen in months. Otherwise, think the end of 2028 at the earliest as things stand. However, as Rollup X's Falcon demo showed, we don't have to wait for mainnet to start making a difference on the ground ahead of then.
Of course, Ethereum's fork pipeline isn't likely to speed up. The more compelling question is how much can get built around it in the meantime, and whether accountability (e.g. via ETHBEAT) keeps pace!
