Ethereum’s scaling roadmap got a detailed public airing this week, straight from Vitalik Buterin himself. The thread covers two distinct timeframes, and the gap between them is wider than most people probably realize.
The Near Term: Glamsterdam Does the Heavy Lifting
The next meaningful upgrade, Glamsterdam, carries more technical weight than its name suggests. Three changes arriving with it are worth understanding together rather than individually.
Block-level access lists allow blocks to be verified in parallel for the first time, directly addressing one of the more stubborn throughput constraints in current execution. ePBS changes how slot time is used, making it safe to dedicate a much larger fraction of each slot to verification rather than just a few hundred milliseconds. Together, these two changes expand what the network can actually process per block without compromising safety.
Now, scaling.
There are two buckets here: short-term and long-term.
Short term scaling I've written about elsewhere. Basically:
* Block level access lists (coming in Glamsterdam) allow blocks to be verified in parallel.
* ePBS (coming in Glamsterdam) has many features, of…— vitalik.eth (@VitalikButerin) February 27, 2026
The third piece is multidimensional gas, and this one is arguably the most structurally significant. Buterin has been working toward separating “state creation” costs from standard execution and calldata costs. Today, a single gas dimension governs everything. Under the new model, writing new state to the chain gets its own gas type with its own cap, which stops state bloat from consuming the same budget as ordinary execution.
Crucially, state creation gas won’t count toward the existing ~16 million transaction gas cap, meaning larger contracts become viable without crowding out regular transactions.
The EVM compatibility challenge here is real. Opcodes like GAS and CALL currently assume one gas dimension. Buterin’s solution introduces a “reservoir” mechanism that maintains backward compatibility while allowing specialized dimensions to be consumed first when available.
Long Term: Blobs and ZK-EVM
Further out, the roadmap rests on two pillars. PeerDAS continues iterating toward roughly 8 MB/sec of blob throughput, enough to serve Ethereum’s L2 ecosystem without trying to become a global data layer. ZK-EVM adoption progresses gradually, eventually enabling full block validation via zero-knowledge proofs without requiring validators to re-execute transactions personally.
Multi-proof validation, where ZK proofs, formal verification, and other methods converge, sits at the far end of that timeline, contingent on security maturity that doesn’t exist yet.
Ethereum’s path forward is technically dense. Buterin’s thread made clear it’s also genuinely well-mapped.
