Nethermind

UBS and Nethermind have completed two joint proofs of concept showing that a public, permissionless network can support the compliance and operational needs of regulated financial institutions. The PoCs show that banks and asset managers can apply strong compliance controls through the infrastructure they run on top of Ethereum, without changing how the protocol itself works. We built and tested a node that applies customizable compliance rules to outgoing transactions, and a routing component that sends approved bundles directly to selected block builders. Both validated end-to-end on Sepolia, no live transactions. This is what enterprise-grade blockchain infrastructure built on deep protocol expertise looks like in practice. We plan to build on this work with @UBS. Full release: ubs.com/global/en/medi…

Confidentiality is a real gap for institutions building on Ethereum, and it's good to see a dedicated team going after it with specs and security properties instead of a deck. We've been doing our own engineering work on the institutional side this past years - privacy is one piece of a larger set of open problems: compliance tooling, custody integration, settlement finality. No single team closes all of that. We're building in this space as well, and where there's overlap on specific technical problems, we're open to that conversation.

If you're an Ethereum RPC or infra provider, execution speed sets your latency floor and memory sets your bill. @Nethermind is the fastest execution client on 90% of Mainnet blocks, most consistent on p95/p99 blocks on <1/3 the peak RAM of the alternative.

And now, 90% of Ethereum blocks over the last 24 hours have been executed fastest by the @Nethermind client.

Great dashboard for tracking Ethereum client performance, created by the EthRex team. @Nethermind has been the fastest client to process approximately 80% of Ethereum blocks (last 12 hours results). Full dashboard: grafana.ethrex.xyz/d/client-cmp-v…

Nethermind 1.39.0 is live: 337 changes since 1.38.1. 𝗥𝗲𝘀𝘁𝗮𝗿𝘁 𝘀𝗮𝗳𝗲𝘁𝘆: Header sync now recovers after a lost chain level on unclean shutdown, and Flat DB writes are durable across restarts, closing a real gap where nodes could come back up corrupted after a bad shutdown. 𝗥𝗼𝗮𝗱 𝘁𝗼 𝗚𝗹𝗮𝗺𝘀𝘁𝗲𝗿𝗱𝗮𝗺: EIP-7928 (Block-level Access Lists) reaches devnet-7 support with a real performance pass, cutting allocations on the BAL generate and validate paths. eth/71 ships alongside it, the wire protocol that lets peers exchange BALs directly. EIP-8037 rounds it out, metering gas for state growth. 𝗝𝗦𝗢𝗡-𝗥𝗣𝗖: Eight new methods, including eth_sendRawTransactionSync, eth_baseFee, and eth_capabilities, plus streaming responses for debug_ and trace_ calls and a correctness pass across eth_estimateGas, eth_simulateV1, and eth_createAccessList. 𝗙𝗶𝘃𝗲 𝗯𝗿𝗲𝗮𝗸𝗶𝗻𝗴 𝗰𝗵𝗮𝗻𝗴𝗲𝘀 𝘁𝗵𝗶𝘀 𝗿𝗲𝗹𝗲𝗮𝘀𝗲: - Legacy BloomStorage removed - Discovery defaults to V4 - Per-method JSON-RPC metrics on by default - Non-spec error field dropped from receipts - IPC sockets restricted to owner-only

Same proof methodology, three independently built Plonky3 codebases. OpenVM and Brevis Pico (EF-supported) are zkVMs. Plonky3's own Poseidon2 circuit isn't, but the method still applies. For the zkVMs, we prove one equivalence theorem per opcode: a chip's circuit, read through its bus interactions, produces exactly the same state as running that RISC-V instruction. Pico alone has 62 of these for RV64IM, 45 for RV32IM. Poseidon2's constraint expressions ran to billions of nodes, too many for the extractor or Lean to handle directly. We cached shared subexpressions down to around 3,000 named terms. On OpenVM, memory consistency, every read and write to an address agreeing with one history, is derived from bus balance rather than assumed. On Pico, one piece stays outside the proof: byte-level lookup tables are audited, not proven in Lean. Running your own Plonky3 fork and weighing formal verification against building this yourself? We're already having that conversation with a few teams.

Disk failure. Hardware refresh. Provider migration. However a node goes down, time offline is time exposed to missed attestations. Nethermind gets back to full sync in 30 minutes, cold start, straight from the Ethereum P2P network. No snapshots, no shortcuts.

Looking back at Nethermind's core development journey over the years makes me incredibly proud of what the team has accomplished. The Resiliency Phase Before the Merge, there was effectively no alternative to Geth. It was by far the most stable and performant execution client, and execution client diversity barely existed. That concentration posted a major risk to Ethereum's resilience and one of its core properties: never going down. Today, Nethermind stands alongside Geth as one of the leading execution clients, with each accounting for roughly 30% of the network. The Ethereum Scaling Phase Once execution client diversity became a reality, the next major challenge was scaling Ethereum. Ethereum had been stuck at around a 30M gas limit for a long time. Nethermind led much of the performance engineering effort, demonstrating that Ethereum could safely operate at significantly higher throughput. Beyond setting an example, many of our optimisations and ideas were adopted by other client teams, helping raise performance across the whole Ethereum. Just as importantly, we helped establish a performance engineering culture within Ethereum core development, bringing a more data-driven approach to gas limits and gas pricing. That work was later adopted and expanded by the EF. Ethereum still has plenty of room to scale, but we're moving in the right direction. What's Next We’re now focused on the next set of challenges: 1. Executing on the Strawmap while leveraging AI as much as possible and exploring how AI agents can make Ethereum development increasingly autonomous. 2. Making Ethereum post-quantum resilient. It's part of the Strawmap, but it's important enough to deserve a separate point. 3. Accelerating institutional adoption of Ethereum. 4. Improving RPC layer - from hardening and design to making it verifiable.

@ethereuminsti Congratulations on the launch. It will be interesting to see how this develops and contributes to institutional adoption. Wishing you success with the initiative.

1/ Announcing Ethereum Institutional An independent non-profit dedicated to accelerating the institutional adoption of Ethereum, its L2s, applications and overall ecosystem.

For a regulated institution, privacy on a public chain only counts if what's hidden from the market stays open to auditors and regulators. Our new privacy series works through that, starting with a shielded pool: private transfers that prove in seconds and settle for under a dollar. Compliance is the next piece.

Own the platform, security, identity, and AI infrastructure a fully remote company across 50+ countries runs on. Nethermind is hiring an IT Department Lead. Reports to the CIO. 5+ yrs leading infra. Remote. jobs.ashbyhq.com/nethermind/bea…

@arcane_finance Privacy for real financial apps on Stellar is hard work, the compliance and disclosure side especially. Good to see Arcane building it. We've been working on this too, built a privacy pool there and are still developing it.

AI agents need economic rails. @tkstanczak of @nethermind explores why autonomous systems operating onchain represent an entirely new asset class and what that means for how we build. Live on the Main Stage at ETHConf.

Great ecosystems are built by great teams. Cohort 1 brings together founders and innovators shaping the future of Ethereum's financial infrastructure. x.com/0xboo/status/2…

@annkkitaaa Welcome to the team!

I'm incredibly excited to share that I'll be joining @NethermindEth as a Research Intern for the next 3 months! Honestly, this is an opportunity I've been hoping and working towards for a long time. From studying mathematics and machine learning to exploring cryptography, Plonky3, Circom, privacy research, and zero-knowledge proofs, every step has only made me more curious about this space. The chance to learn from and work with researchers and engineers building at the frontier of Web3 and cryptography is something I'm genuinely grateful for. Here's to three months of learning, research, building, and asking way too many questions. Let's go! 🔥 @Nethermind

ETHConf, next week. The institutions come to Ethereum: SEC on tokenized securities, DTCC on settlement, BlackRock on why they want to be onchain. It all settles on the execution layer. We've been building that layer since 2017. In NY all week. Come talk infrastructure.

Release notes: github.com/NethermindEth/…

Nethermind 1.38.0 is live: 51 contributors. 372 changes. Nethermind co-authored EIP-7928 (Block-level Access Lists), which lands in this release. BALs let the client declare storage accesses up front and execute independent transactions in parallel rather than sequentially. EIP-7976 increases calldata floor cost and EIP-7981 prices access list tokens. All three are implemented ahead of Glamsterdam activation. Several performance improvements: Hot-path optimizations across the execution layer. Address bytes are now embedded inline, removing a secondary allocation. EVM stack and EvmMemory tighten, parallel execution reduces worldstate and txprocessor allocations, and precompile caching extends. EraE support arrives. Operators can prune historical data from local storage and re-import any block range from a SHA-256-verified remote archive, covering pre and post-merge. The P2P layer rewrite cuts per-peer allocation by 54MB and 2.25M objects. Flat DB now persists its layout and refuses to start under a mismatched one, eliminating a class of restart-time corruption.