World Chain EIP-7928 Targets Parallel L2 Validation

Overview

The World Chain EIP-7928 implementation is designed to stream block-level access lists through Flashblocks, allowing validators to begin parallel verification while a block is still being constructed. World Chain enabled the system on Sepolia on July 27, 2026, and scheduled its mainnet activation for August 17. It therefore should not yet be described as fully active on mainnet.

EIP-7928 records the accounts, storage locations, code changes, and resulting values touched during block execution. By knowing those dependencies in advance, a verifier can process independent transactions across multiple CPU cores, prefetch state from storage, and calculate the state root in parallel. World Chain extends the proposed Ethereum design by transmitting a slice of the access list with each 200-millisecond Flashblock rather than waiting for a completed block.

Internal test-network benchmarks reached up to 1 gigagas per second without validation latency rising proportionally. These results are technically significant, but they remain controlled benchmarks. Mainnet performance, bandwidth costs, failure handling, and decentralization effects will determine whether the architecture delivers comparable benefits under real workloads.

Key Takeaways

• World Chain Sepolia enabled streamed block access lists on July 27; mainnet activation is scheduled for August 17.
• Each 200-millisecond Flashblock carries part of an EIP-7928 access list.
• Validators can execute independent transactions and compute state changes in parallel.
• World Chain reported testing up to 1 gigagas per second on standard cloud infrastructure.
• Benchmark throughput is not the same as sustained mainnet capacity or user demand.

How World Chain EIP-7928 Works

What problem does World Chain EIP-7928 address?

It addresses the sequential verification bottleneck created when validators do not know which parts of blockchain state a transaction will access. Under the conventional model, a verifier often executes transactions in order because a later transaction may depend on a balance or storage value changed by an earlier one.

Block-level access lists make these dependencies explicit. EIP-7928 defines a structure that records the accounts and storage slots accessed during block execution, together with post-execution values. Validators can identify transactions that do not depend on one another and process them across multiple workers.

The same data can support parallel disk reads, transaction validation, state-root calculation, and executionless state updates. A list must match the actual execution trace; inaccurate results can cause the block to be rejected. The formal EIP-7928 specification therefore treats the access list as a verifiable component rather than an optional performance hint.

How are access lists delivered through Flashblocks?

World Chain produces a block every two seconds and divides the work-in-progress block into signed Flashblocks delivered approximately every 200 milliseconds. Each Flashblock carries the transactions added during that interval and a corresponding slice of the block access list.

The slices contain minimum and maximum transaction indexes so that validators can assemble them into a complete list for the block. Each Flashblock also includes an access_list_hash, creating a cryptographic commitment to the state accessed and changed during that portion of execution.

Followers can begin verification as soon as each slice arrives. Transactions are distributed across a worker pool, results are reassembled in index order, and the claimed state changes are checked against execution. State-root calculation can proceed in the background. By the time the final Flashblock arrives, much of the verification work has already been completed.

Why Streamed Access Lists Could Improve Scaling

Can the design increase throughput without stronger hardware?

World Chain’s tests suggest it can improve verification throughput without proportionally increasing verifier hardware, but mainnet evidence is still required. The project tested the design on standard AWS cloud infrastructure with the execution client limited to four virtual CPUs and ordinary EBS gp3 storage.

A nine-point test increased sustained throughput from 1.7 Mgas/s to 1 Ggas/s against a two-billion-gas block limit. World Chain reported that validation latency increased at lower throughput, reached a knee near 100 Mgas/s, and then broadly plateaued as work was distributed across CPU workers and Flashblock intervals.

At saturation, the stated constraints were the load generator and sequencer rather than the validator. This indicates that parallel verification may weaken the traditional relationship between larger blocks and higher end-of-block validation latency. The results and testing environment are detailed in World Chain’s engineering report.

The limitation is that these were project-run tests in a development cluster. Mainnet transactions may produce different dependency patterns, database behavior, network conditions, and state growth.

How does World Chain differ from Ethereum’s planned design?

Ethereum has selected EIP-7928 as a headline component of the planned Glamsterdam upgrade. Its proposed Layer-1 implementation uses one block access list for a roughly 12-second block, places a commitment in the block header, and requires activation through a coordinated hard fork.

World Chain uses the same underlying concept but adapts it to a faster Layer-2 pipeline. Instead of one list per block, it distributes one access-list slice per 200-millisecond Flashblock. Commitments appear in signed Flashblock deltas, and the data travels through World Chain’s peer-to-peer protocol.

World Chain can enable the feature through a runtime flag, so the rollout does not require a coordinated network hard fork. Sepolia activation occurred on July 27, while mainnet activation is scheduled for August 17. The implementation makes World Chain an early production-oriented test of an execution concept that Ethereum plans to adopt later.

Security, Bandwidth, and Operational Limits

Can a dishonest access list make an invalid block valid?

No, provided validators independently execute and compare the results, but a dishonest list may still waste validation resources. The block builder produces the access data during execution, while followers verify that the declared reads, writes, and post-values match their own results.

If the list omits required state or reports incorrect changes, the discrepancy should be detected and the block rejected. Signed hashes also commit the builder to each streamed slice, allowing divergence to be identified within a Flashblock interval rather than after the entire two-second block.

This model does introduce implementation complexity. Clients must correctly assemble slices, enforce transaction-index boundaries, manage parallel execution, and avoid race conditions. Faults in the parallelization or commitment logic could become consensus or availability risks even when the theoretical design is sound.

What should be measured after the August 17 rollout?

The main indicators should include sustained gas throughput, per-Flashblock validation latency, CPU utilization, database input/output, bandwidth consumption, state growth, and follower synchronization performance. World Chain measured approximately 2–3 KiB of access-list data per Mgas under deliberately storage-heavy tests, while compute-intensive workloads generated proportionally smaller lists.

Operational stability matters as much as peak throughput. Observers should track missed blocks, sequencer interruptions, reorganization behavior, verification failures, and the performance of nodes with different hardware configurations.

Demand is another constraint. Technical capacity does not guarantee that users or applications will consume it. Higher limits may benefit high-frequency payments, games, social applications, and complex onchain markets, but real adoption must be separated from benchmark capacity.

Mainnet Data Must Validate World Chain’s Scaling Claim

World Chain EIP-7928 is an important attempt to change how Layer-2 throughput is verified. Instead of completing a large block and then requiring followers to replay its transactions sequentially, the design communicates state dependencies throughout block construction. This allows verification, state prefetching, and state-root calculation to occur in parallel and across the full block interval.

The approach is especially relevant because increasing block capacity without changing verification architecture can raise node requirements and weaken independent participation. World Chain’s benchmarks indicate that streamed access lists may provide substantially higher throughput while keeping validator hardware relatively modest. The runtime-flag rollout also gives the network a faster deployment path than Ethereum Layer 1, where the equivalent change requires consensus-wide hard-fork coordination.

Still, the strongest claims remain provisional until the August 17 mainnet activation produces measurable results. Controlled tests cannot fully represent mainnet state access patterns, network variability, client failures, or adversarial behavior. World Chain EIP-7928 should therefore be viewed as a significant production experiment in parallel blockchain verification—not yet proof that one-gigagas throughput can be sustained safely under real demand.

Sources

https://eips.ethereum.org/EIPS/eip-7928

https://world.org/blog/engineering/world-chain-full-block-access-lists

Risk Disclaimer: This article is for reference only and does not constitute investment advice. The cryptocurrency market is highly volatile. Please make decisions cautiously based on your individual circumstances.