Ethereum Pectra upgrade and L2 sequencer decentralization reality

EIP-7691 increases the target number of blobs per block from 3 to 6 and the maximum to 9, which encourages rollups to move away from calldata and toward the cheaper blob space. EIP-7623 caps the size of blocks containing excessive calldata to prevent bandwidth overload. Rollups use these blobs because they store data in the consensus layer instead of the execution layer. This separation keeps gas costs low since smart contracts cannot access the blob bytes directly. Developers use the KZG commitment to verify data without the full payload. EIP-7523 helps by clearing empty accounts and restricting unnecessary contract creation to reduce state bloat. Pectra also introduced EIP-7702, which allows wallets to act like smart contracts for single transactions.

The KZG commitment compresses a polynomial of degree 4,095 into a single point on an elliptic curve. This allows anyone to prove a specific value belongs to a polynomial without revealing the rest.

Centralized sequencers remain the norm

The promise of decentralization lags behind the throughput improvements. Every major Layer 2 rollup currently relies on a single operator to order transactions and build blocks. Arbitrum uses Offchain Labs, Base uses Coinbase, and Optimism uses the Optimism Foundation. This centralization creates risks including censorship, MEV extraction, and liveness failure. If a sequencer goes offline, the entire chain stops. I would ignore the marketing claims about decentralization for now because the actual production-grade transition for most major rollups will not happen until late 2026 or 2027.

A sequencer performs four functions: transaction ordering, block building, batch posting, and state root proposal. The tension between speed and decentralization remains evident because centralized sequencers achieve sub-second confirmation which users expect, while a multi-operator system introduces consensus latency that could regress the user experience. Arbitrum launched timeboost to auction MEV, yet full multi-party sequencing remains a target for late 2026. Linea paused its sequencer for one hour in June 2024 following a DEX exploit. Base also experienced a downtime event in February 2025. Espresso Systems operates Mainnet 0 with a permissioned set of 100 nodes using HotShot consensus. Astria, a competing shared sequencer effort, shut down in December 2025.

Will the industry finally move away from single-operator control?

Efficiency gains and market consolidation

Efficiency gains from EIP-4844 and Pectra drive users toward dominant platforms. Base, Arbitrum, and Optimism process nearly 90% of all Layer 2 transactions. Base captured 60% of activity, while Arbitrum maintains over $18 billion in total value locked. This concentration makes sense because users prefer liquidity and proven security over a fragmented ecosystem of hundreds of small chains. The total economic value secured by rollup solutions surpassed 40 billion dollars as of April 2024.

You should watch how these winners manage their own risk. Base reached Stage 1 in April 2025 with permissionless fault proofs, but Coinbase still operates the only sequencer. Optimism targets 2026 for integrated shared sequencing with Espresso and Flashbots. Arbitrum moves toward multi-party sequencing with shared revenue. Most other rollups, including Linea and Scroll, lack a confirmed mainnet date for decentralization. Arbitrum and zkSync Era achieved significant reductions in user delay times, demonstrating improved efficiency in these cases.

Rollup costs dropped from daily averages of $888.6k before Dencun to $135.1k after. Optimistic rollup margins grew from 22.65% to 92.3% in the 150 days following the implementation of EIP-4844. The Base rollup experienced a 224% rise in transaction volume following the implementation of EIP-4844.