L2 INTEL / An ongoing report on Layer-2 blockchain infrastructure

VOL.02
ISSUE04
EDITIONPUBLIC
EDITORM. ARELLANO
WORDS4,812
§ LEDE

Layer-2s do not arrive in the world the way new chains do. They arrive the way new newspapers arrive — quietly, with a masthead, a column inch budget, and a small team of editors arguing about what counts as a fact. Volume Two follows the people who run those rollups, the engineers who write their fraud proofs, and the analysts who decide whether a sequencer outage is a footnote or an obituary. We started this issue in March, finished it on a Sunday, and we are still learning.

What follows is split deliberately: numbers on one side, language on the other, the seam down the middle a reminder that the two halves of the rollup story rarely agree. Read across, not down. The story is in the gap.

FILED BY M. ARELLANO & STAFF FROM REMOTE

EXHIBIT Sequencer cabinet, secondary site. Single rust-coloured indicator marks a witness node currently offline for upgrade.
§ DATA  /  EXHIBIT A

The four rollups we keep on file

A snapshot taken on a Tuesday morning. Numbers move. The story underneath them moves more slowly.

L2 TVL (M USD) TPS FINALITY SEQS
Arbitrum One 2,184 11.4 7m 12s 1
Optimism 1,041 8.7 7m 03s 1
Base 1,612 14.2 6m 48s 1
zkSync Era 486 9.1 7m 52s 1

METHOD seven-day rolling means; finality is median time to L1 confirmation; values rounded to the nearest unit reported by each project's public dashboard. We do not adjudicate disputes.

FIG. 01

Throughput, last seven days

TPS, peak hourly
051015 TPS

Cyan bar marks the metric our analyst desk is currently watching. Other bars decorate but do not lead.

§ ANALYSIS
PULL QUOTE
“The interesting question is not how fast a rollup is. It is who gets to decide that the rollup stopped.”
D. KOVAC
Independent rollup researcher

Sequencer governance is a slower story than throughput

Last week three of the four rollups we follow disclosed pieces of their governance roadmap. Read together, the disclosures argue that the next year of this beat is about handoff — from operator to committee, from committee to chain.

Optimism's OP-117 is a draft, not a vote. It proposes a six-of-eight signer set drawn from independent operators, a public timeout window, and a forced-inclusion path that does not require a rollup-team relay. The text is plain. The implications are not. A six-of-eight committee with disclosed identities is a more public document than any of the existing rollup teams have ever signed.

Arbitrum's response, which arrived two days later, was to publish its current single-sequencer SLA in unprecedented detail — uptime, batch cadence, and the exact list of incidents in the last quarter. The document is dry on purpose. We read it as a defence of staying single-sequencer for one more cycle while the BOLD fraud-proof system stabilises.

Base, meanwhile, said almost nothing. The silence is itself a position. Coinbase is running the only meaningful single-operator rollup whose operator has a public balance sheet, and the lack of a roadmap for sequencer decentralisation is now the loudest disclosure in the set.

ANALYSIS BY M. ARELLANO

§ ARCHIVE

Past issues, on file

Each prior issue lives at its own URL and remains corrigible. Errata are appended, never silently fixed.