World Chain Ships EIP-7928 Access Lists Before Ethereum

On August 17, 2026, World Chain became the first production Layer 2 network to stream full EIP-7928 Block Access Lists, publishing them roughly every 200 milliseconds inside its flashblocks. Ethereum itself has EIP-7928 slated for a later upgrade, so this is a real network running a proposed mainnet feature ahead of the base layer. For anyone tracking how blockchains get faster without getting more expensive to run, it is the most instructive deployment of the month.

• World Chain activated streamed EIP-7928 Block Access Lists on mainnet on August 17, 2026
• Access lists publish roughly every 200ms inside each flashblock rather than once per completed block
• The stated throughput target is up to 1 gigagas per second with no increase in validator hardware requirements
• Activation used a runtime flag, so client operators upgraded without a coordinated hard fork

What Is EIP-7928 and Why Does It Speed Things Up?

EIP-7928 introduces Block-Level Access Lists: a consensus-verified record of every account and storage slot a block touches while executing. Today a validator largely discovers those dependencies as it goes, which forces it to process transactions in order because it cannot know in advance which ones might collide.

Publish the access list and that changes. A validator can see which transactions are genuinely independent and verify them simultaneously across multiple CPU cores. The performance gain does not come from faster hardware; it comes from finally being allowed to use the cores that were already sitting there. That is why the throughput target of up to 1 gigagas per second is paired with the claim that validator hardware requirements do not rise.

Why Streaming the List Every 200 Milliseconds Matters

This is World Chain's specific contribution. Rather than emitting one access list when a block is finished, the network streams the list continuously inside flashblocks, its sub-block units produced roughly every 200 milliseconds.

The practical effect is that verification starts while the block is still being assembled instead of afterwards. Validation work spreads across the whole block-building window rather than piling up at the end. Reporting on the rollout notes that validation latency remained effectively stable even as throughput rose, which is the outcome you want: more capacity that does not translate into slower confirmation or heavier machines.

How Did It Ship Without a Hard Fork?

By runtime flag. Client operators could enable the feature independently rather than coordinating a synchronised network upgrade, which is why World Chain could move first. Ethereum plans to adopt EIP-7928 as part of its Glamsterdam upgrade, and a Layer 2 running it in production first gives core developers something better than a testnet: real traffic, real block builders, real edge cases.

This is becoming the normal division of labour. Layer 2s iterate on execution improvements quickly, and the base layer adopts what survives contact with production. We saw the client-side version of the same story when Solana's Agave 4.2 cut slot times to 350ms, and the roadmap side when Ethereum's Hegota upgrade shortlisted 66 EIPs.

What to Watch Next

The number that will settle the argument is sustained throughput under real load, not the headline target. One gigagas per second is a design goal; what matters is what the network holds during peak activity while validation latency stays flat and independent node operators keep running on ordinary hardware. Watch also for other Layer 2s adopting streamed access lists, because a single deployment is a demonstration and three or four is a standard.

For more on protocol upgrades and scaling work, follow our crypto coverage.

Sources: The Block — August 5, 2026; Bitcoin.com News — August 2026; Invezz — August 5, 2026.