Vitalik trong đợt "cải cách tối giản" này rốt cuộc muốn giải quyết điều gì?

Author: William M. Peaster | Compiled by: BAIHUA Blockchain

Ethereum research circles have been on fire lately. From July 4 to now, Vitalik Buterin first reorganized Ethereum's long-term direction, then published a new research post, "The Extremely Lean Chain", and specifically highlighted a proposal to introduce Bitcoin-style UTXOs into Ethereum.

These developments are interesting on their own, but when viewed together, they offer a clearer picture of what Ethereum is evolving into.

Two weeks ago, Ethereum researchers met in Berlin to continue discussions with client teams that began in Svalbard in April, planning the protocol's long-term evolution roadmap. The updated strawmap has been released, and I've attached the image in this post. — vitalik.eth, July 4, 2026

The context here is the so-called strawmap, which the Ethereum Foundation refers to as a "rough draft roadmap"—a sketch of L1 upgrades covering the rest of this decade, last updated in late June. Then on July 4, Vitalik gave his own summary of these changes:

"Lean Ethereum is not a single upgrade completed all at once, but a set of improvements that will roll out to the Ethereum network over the next three to four years. But don't be mistaken: it is indeed the third major iteration of Ethereum, just as The Merge was the second."

He went on to say that along this evolution arc, nearly all major components of the network will be replaced: how blocks are validated, how consensus is reached, and even what "state" itself means. So what happens next?

Of course, the strawmap is not a single narrative thread. But from these major trends, combined with Vitalik's commentary, we can already distill several big themes. For example:

Validation is replacing re-execution — Currently, every Ethereum node re-executes every transaction to check that on-chain computations are correct. In the Lean era, nodes will instead verify cryptographic proofs, i.e., recursive STARKs. This change will make "proving correctness" itself cheaper, further driving execution scaling, lowering hardware requirements, and opening up more optimization opportunities.

Ethereum is going on a "state diet" — A multi-layer state system is taking shape. Today's flexible but heavier "dynamic" state will remain, but its future growth will be more limited. Conversely, several new types of state will be cheaper and less flexible, but will be expanded more aggressively. Vitalik gave a hypothetical 2030 Ethereum: the former about 2TB, the latter potentially expanding to 100TB. Migration to these new state types will not be mandatory, but economic incentives will be in play because they offer significantly lower fees for projects and users.

Privacy and quantum resistance become design pillars — While most public blockchains are still 1) completely transparent, and 2) slow to act on quantum resistance, Ethereum researchers have elevated privacy UX and quantum defense to core design principles that must be considered upfront and built around. As Vitalik said: "When designing Frames, mempool, and state tree additions, we explicitly ask: 'How can a quantum-secure, intermediary-independent privacy protocol transaction pass through this system? What is its cost?'"

These advances will not land in a single upgrade like The Merge, but will be spread across 6 to 7 forks between now and 2029. Of course, it's called a strawmap precisely because it's only a rough roadmap; the Ethereum Foundation has always emphasized that this timeline and route are more of a coordination reference than a finalized construction blueprint.

However, it's exciting that new mechanism proposals are now appearing almost daily.

A current example: this morning, Vitalik released "The Extremely Lean Chain", a design proposal aiming to compress Ethereum's consensus layer to near-ultra-lean status.

In the current Ethereum architecture, the Beacon Chain needs to store a rather bulky record for each validator, and processes balance updates for all validators every epoch. In Vitalik's evolved version, the on-chain data stored per validator is only about 6 bytes, down from approximately 121 bytes today—a reduction of 95%.

The mechanism works as follows: each day, every ETH staker generates a ZK proof of their updated balance and submits it on-chain. This way, Ethereum essentially shifts from "doing the accounting itself" to "checking receipts." If a staker misses a day's proof submission, they simply cannot participate in attestation temporarily; once they catch up, they can resume—no slashing involved.

As Vitalik put it, the biggest benefit of this path is that it "could allow the consensus layer to scale to millions of validators if needed." Could this lay the groundwork for lowering the 32 ETH staking threshold in the future? We'll see. But the design has other advantages too, such as in its full form, validators could even re-register with a new public key every day, paving the way for anonymized staking.

However, Vitalik's "Extremely Lean" proposal is just one direction worth noting right now. Another mechanism draft that could equally have a huge impact on Ethereum was just released by Ethereum Foundation researcher Toni Wahrstätter, titled "Native UTXOs on Ethereum."

What if we simply borrow Bitcoin's transaction model? That's the core of this proposal. Today on Ethereum, receiving a payment leaves a permanent record. When an address first holds ETH or some token, every node must store that state item forever—even if the address is only used once. Multiply that by billions of payments, and state bloat becomes a real crisis.

Wahrstätter's idea is to shift to Bitcoin-style UTXOs, i.e., "unspent transaction outputs"—essentially one-time value packages that get consumed when spent. More specifically, in the Ethereum version, the chain doesn't even need to store the UTXO itself. Its details can remain in the chain's historical data and be proven when needed; the chain's permanent state only needs to keep a single bit per UTXO to mark whether it has been spent.

According to Wahrstätter's estimates, this shift would reduce the permanent state associated with payment flows by about 99.8%. Combined with innovations like frame transactions, this new transaction type could also allow newly generated addresses to receive and then spend funds without holding ETH for gas, paving the way for smoother stealth address experiences on L1.

Zooming out: if concepts like these eventually land on Ethereum, the network will become healthier, more durable, more flexible, and closer to that forward-looking North Star. Personally, I'm most interested in these new state types and how they will affect the application layer, fungible tokens, and NFT experiments.

But regardless of what the next step is and when it comes, Ethereum's direction is now clearer than ever. That's bullish.