Ethereum's Blockchain Rethink: Vitalik Says Layer 2s Must Change as Fusaka Nears December 3

30/10/2026 7 Min. Lesezeit

Aufrufe

Ethereum is quietly rewriting its own scaling story, and the rewrite comes from the top. Co-founder Vitalik Buterin has told the ecosystem that the original rollup-centric roadmap no longer reflects reality: Ethereum's base layer is scaling directly, Layer-2 networks are no longer essential simply as capacity providers, and their future value has to come from somewhere else. The message landed on October 29, just weeks before the network's next major upgrade, Fusaka, is due to activate on mainnet on December 3. For a $410 billion blockchain whose Layer-2 chains carry a $47 billion on-chain economy, that is not a technical footnote. It is a change of direction, and it reframes what every Layer-2 project — Arbitrum, Optimism, Base and the rest — now has to prove.

Start with why the old story made sense. Ethereum's base layer was designed for security and decentralisation, not for raw throughput: its current maximum is around 30 transactions per second, while developers have talked about a long-term goal of 100,000 transactions per second across the network. Layer-2 chains were the workaround — extra lanes on the highway that bundle thousands of transactions, submit summaries called blobs to the base layer, and inherit Ethereum's security while offering faster, cheaper transactions. The model worked, and it built a $47 billion economy on top of Ethereum. But it also created a dependency: if the base layer could not keep up with the data those lanes produced, the whole system would jam.

Fusaka is the next attempt to unjam it. The upgrade package contains 12 changes, and its centrepiece is PeerDAS — Peer Data Availability Sampling. Today, verifying Layer-2 transactions requires nodes to handle the full blob data; PeerDAS lets nodes check only random samples within that data to verify it, sharply reducing the data load. Fusaka also doubles the maximum smart-contract size from 24KB to 48KB. One week after Fusaka, developers plan the first of several "Blob Parameter Only" updates, raising blob capacity from the current 6 to 9 blobs per block to 10–15, then to 14–21 a week later — more than doubling blob capacity within two weeks of activation. The precedent is encouraging: when the Dencun upgrade introduced blobs in March 2024, some Layer-2 networks saw monthly costs fall from $15,000 to $1,500. Researchers estimate these improvements could help Ethereum process up to 12,000 transactions per second across all its layers — still far from the 100,000 long-term goal, but a very different network from the 30-transactions-per-second base layer of today.

That is exactly why Buterin's rethink matters. His argument, reported on October 29, is that gas-limit increases and protocol upgrades have already expanded Layer-1 throughput and cut fees, making direct mainnet usage more attractive — monthly active addresses on Ethereum's Layer-1 are rising sharply even as aggregate Layer-2 usage has declined. In his words, the vision of Layer-2s as "branded shards" of Ethereum "no longer makes sense." Many popular Layer-2s, he noted, are not fully decentralised either: Arbitrum, for example, uses a centralised sequencer controlled by its developer team, Offchain Labs. If a Layer-2 uses ETH or Ethereum-issued assets, Buterin argues, it should reach at least "stage one" decentralisation — limited training wheels — or it is effectively a separate Layer-1 that should use a bridge.

So what are Layer-2s for, if not cheap capacity? Buterin's answer is a spectrum, not a shutdown. Rather than abandoning Layer-2s, he reframes them as places that offer something the base layer does not: privacy-focused virtual machines, application-specific efficiency, ultra-low-latency sequencing, non-financial designs for social or identity use cases, and even built-in oracles or dispute resolution. On the protocol side, he highlighted progress toward native rollups — a proposed rollup precompile that would let Ethereum verify zero-knowledge EVM proofs at the protocol level, with upgrades and bug fixes handled through normal network upgrades instead of external governance. Competing solely on lower fees, his message runs, is no longer enough — and the market seems to agree: ether has plunged nearly 30 percent over the past month, is down 54 percent from its all-time high, and traded around $2,267, down 2.4 percent in 24 hours, in the October 29 coverage.

There is a counter-signal worth noting, because it complicates the bearish price story. The average cost of an ETH transfer has dropped to around $0.095 from this year's peak of $0.72 on April 21, according to Santiment — an 85 percent-plus collapse in fees — as Fusaka, higher blob throughput and a 60 million gas limit have expanded capacity, while Layer-2s process transactions that previously competed for mainnet blockspace. Spot Ethereum ETFs, meanwhile, recorded a net inflow of about $14 million, led by BlackRock's ETHA fund, marking a reversal after several days of outflows. Cheap transactions do not necessarily mean demand is recovering, as Santiment itself cautions — but they do mean the network Buterin is describing is already arriving: a base layer that is cheaper, roomier and more directly usable than the one the rollup roadmap was built for.

What to watch next, in order. First, Fusaka's mainnet activation on December 3, and whether the blob-parameter updates that follow actually deliver the promised capacity without incident. Second, whether Layer-2 teams respond to Buterin's challenge by shipping genuinely different features — privacy, latency, application-specific design — rather than another round of fee cuts. Third, whether the native-rollup precompile moves from proposal to protocol, because that would change the governance bargain for every existing rollup. And fourth, whether mainnet activity keeps rising while Layer-2 usage falls, or whether the two find a new equilibrium once Fusaka lands. Ethereum's scaling debate used to be about how to build more lanes. After October 29, it is about what the lanes are for — and December 3 is when the new answer starts getting tested in production.

FAQ:

What is Fusaka, and when does it activate? Ethereum's next major upgrade package, due on mainnet on December 3, containing 12 changes headlined by PeerDAS data-availability sampling.

What does PeerDAS do? It lets nodes verify Layer-2 blob data by checking random samples instead of downloading the whole blob.

How much capacity does Fusaka add? Blob capacity rises from 6–9 blobs per block to 10–15, then 14–21 a week later; researchers estimate up to 12,000 transactions per second across all Ethereum layers.

What did Vitalik Buterin say about Layer-2s? That the rollup-centric roadmap no longer reflects reality, that Layer-2s acting as "branded shards" "no longer makes sense," and that their future value must come from features like privacy, latency or application-specific design.

What is a native rollup? A proposed rollup precompile letting Ethereum verify zero-knowledge EVM proofs at the protocol level.

How cheap has Ethereum become? Average ETH transfer about $0.095, down from a $0.72 peak on April 21 (Santiment).

What is ether's price context? Around $2,267, down 2.4 percent in 24 hours, nearly 30 percent over the past month, 54 percent from its all-time high; spot ETH ETFs saw about $14 million net inflows led by BlackRock's ETHA.

Is this the same as the Glamsterdam blog? No — Glamsterdam was a separate standalone story; this blog is the October 29 Vitalik Layer-2 rethink plus December 3 Fusaka/PeerDAS.