Back to all reports

以太坊研究日报

Ethereum Research Daily

Vitalik wrote a note on the recursive STARK mempool (EIP-8288), which he hopes gets into I*, the fork after Hegotá, as the next step after frame transactions; his EIP for a frame type for PQ signature and STARK aggregation was merged as a draft the same day. misilva73 simulated the first day of the gas limit rising from 60M to 200M: fees spike and then fall, and 9 of 12 demand scenarios end with median utilization near the target.

Image for September 9, 2026

Image: GitHub, ethereum/EIPs PR #11772

After frames: recursive STARK aggregation

EIP-8288 update

New EIP: frame type for PQ signature and STARK aggregation (vbuterin) was merged. Transactions declare signatures and STARKs (for privacy or for L2s) as "dependencies", and the mempool and block builder can replace them with a recursive STARK proving they all exist, supporting them very gas-efficiently in a post-quantum world. An EIP-8288 update was also merged.

https://github.com/ethereum/EIPs/pull/12308

Glamsterdam's gas capacity

ethresear.ch: from 60M to 200M, simulating Glamsterdam's fee market

14:43 (misilva73). Around 21 days of historical mainnet transactions were replayed under the new gas schedule to simulate the first 7,200 blocks (about a day) under 12 demand scenarios. The limit takes about 4 hours to reach 200M; the day shows an initial fee spike, a decline as capacity expands, then adjustment around the new 100M target. Nine scenarios end with median utilization near target and median base fees from 0.0002 to 0.3928 gwei; the other three stay below target even with a nearly negligible base fee.

https://ethresear.ch/t/from-60m-to-200m-simulating-glamsterdam-s-fee-market/25957

Other updates

Community