Core Engineering for Ethereum, Rollups & ZK

Execution Infrastructure for Live Networks

The execution layer underpins Ethereum transaction execution and rollup settlement. These systems depend on execution infrastructure operating correctly under sustained network load. The Core Engineering team builds and operates execution infrastructure across Ethereum mainnet, Layer 2s, zk execution environments, and modular execution stacks. Maintaining an execution client at production scale surfaces problems that only appear under live network pressure. Core Engineering contributes directly to Ethereum and adjacent execution ecosystems, with engineering work aligned closely to protocol evolution, scalability research, and long-term infrastructure requirements. The same engineering work now extends into rollup sequencing and zkVM-compatible execution.

PROBLEMS OF live network

Worst-case execution paths

History growth

Validator timing constraints

RPC reliability under sustained query load

Cross-client protocol coordination during upgrades

Engineering domains

Six domains of execution engineering, all running in production and held to one standard.

The Nethermind Ethereum Client

An Ethereum execution client deployed across Ethereum mainnet, rollups, staking infrastructure, and enterprise deployments.

Client Reliability and Release Engineering

Test coverage has to grow with the codebase, and the node fleet has to mirror mainnet. Over 1.4 million test executions run per commit and over 500 million per release, across a fleet of more than 20 production and devnet nodes.

Ethereum Protocol Engineering

Direct participation in Ethereum's execution-layer roadmap. Nethermind engineers co-authored 7 of the 12 EIPs included in Fusaka, covering gas-limit expansion, execution-layer scalability, rollup economics, proposer predictability, blob fee mechanics, and execution performance safeguards.

Layer 2 & Rollup Infrastructure

Production engineering across optimistic, ZK, and based rollup architectures. Work includes execution clients, sequencing systems, prover integrations, benchmarking, developer tooling, and protocol optimization.

ZK Execution Environments

Nethermind's Ethereum execution semantics compiled to RISC-V64 and integrated across multiple zkVM environments. The work spans execution witness generation, stateless block replay, and zkVM-compatible execution binaries. This is the engineering required to make Ethereum execution ZK-provable as those environments mature.

Performance Engineering & Infrastructure Research

The Gas Benchmarking Framework developed by Nethermind is now used across Ethereum client teams and Ethereum Foundation research groups to validate readiness for gas-limit expansion. Related work covers worst-case execution behavior, throughput validation, execution parallelization research, and adversarial performance testing.

Technical approach

Core Engineering combines execution-client work, protocol research, low-level programming optimization, benchmarking infrastructure, and ZK systems engineering.

The team also delivers bespoke client implementations for chains with non-standard requirements, including Gnosis, XDC, and several rollup ecosystems.

Core technologies

.NET runtime

FlatDB

RISC-V64

Zisk

Ethereum execution semantics

Engineering challenges and failure modes

Execution-layer engineering operates under a small set of recurring constraints. These are the failure modes the team's work is built to defend against.

exposure

Execution-client bugs or behavioral inconsistencies can cause nodes to diverge from network consensus, creating forks.

defense

The team's testing infrastructure and quality assurance procedures are built against this failure mode directly.

exposure

Execution clients without sufficient performance headroom can collapse under adversarial query volume or transaction surges.

defense

This is a primary motivation for the team's performance and gas benchmarking work.

exposure

As Ethereum's history accumulates and gas limits expand, execution paths and storage layers become bottlenecks.

defense

Nethermind built FlatDB to cut trie traversal during execution. The team's EIP-4444 history expiry implementation, with Era1 and EraE import/export formats, addresses history growth directly.

exposure

Each Ethereum upgrade introduces coordination requirements across client teams.

defense

Nethermind engineers contribute to EIP authorship and cross-client testing to reduce upgrade risk.

exposure

Non-standard chains carry real maintenance overhead.

defense

Nethermind built FlatDB to cut trie traversal during execution. The team's EIP-4444 history expiry implementation, with Era1 and EraE import/export formats, addresses history growth directly.

When teams engage core engineering

Network throughput at current limits

Gas-limit expansion, execution bottlenecks, and history growth surface as throughput pressure. The team's work on the Gas Benchmarking Framework, FlatDB, and history growth handling addresses these directly.

RPC reliability under load

When query volume degrades response times or surfaces inconsistent behavior, the work moves to endpoint standardization and RPC performance. The team's recent eth_getLogs scaling work is one example, with broader RPC performance gains in progress.

Rollup sequencing and preconfirmation latency

The sequencing layer sets preconfirmation latency on L2s. The team builds and runs that infrastructure for rollups. Catalyst operates Taiko sequencing in production.

Synchronous composability across rollup ecosystems

Fragmented liquidity between L2s is addressed through execution architectures that maintain composability with mainnet. The team's engineering experience in this area spans earlier production deployments and ongoing work.

Client features for non-standard chains

Chains like Gnosis and XDC require client-side features outside the standard Ethereum execution model. Core Engineering builds and maintains these as part of the client.

Alternative clients for adjacent networks

Juno is Nethermind's independent Starknet full-node client. Pluto brings client diversity to Obol's Distributed Validator middleware. Voyager is the Nethermind-built block explorer for Starknet.

Selected engineering work

A selection of production systems, protocol contributions, and research tooling built and maintained by the Core Engineering team.

Nethermind Ethereum Client

Nethermind's .NET-based Ethereum execution client, built for high-performance syncing, modular plugin architecture, and extensibility across new chains and features.

Ethereum

•

Staking

•

Enterprise

Fusaka EIP Authorship

Nethermind engineers co-authored 7 of the 12 EIPs included in Ethereum's Fusaka upgrade, covering gas-limit expansion, rollup economics, blob fee mechanics, and execution performance safeguards.

Ethereum Foundation

•

Fusaka

Gnosis Chain Merge

The sole execution client supporting Gnosis Chain through its transition from Aura to Proof of Stake, and one of the longest-running production deployments of the Nethermind client outside Ethereum mainnet.

Gnosis Chain

Catalyst

A preconfirmation sidecar engineered for Taiko Alethia, currently operating in production as the sequencing layer for Taiko's based rollup. It cuts transaction confirmation time from over 20 seconds to roughly 1 second.

Taiko

•

Live in production

Juno & Voyager

Independent Starknet full-node implementation (Juno) and block explorer (Voyager), built and maintained by Nethermind.

Starknet

Pluto

An alternative Distributed Validator middleware client built by Nethermind for the Obol Network, bringing client diversity to Ethereum's DV layer alongside Obol's Charon.

Ethereum

•

DVT

•

Obol Network

Gas Benchmarking Framework

Originated by Nethermind and now used across Ethereum client and Ethereum Foundation research teams to validate execution readiness for future gas-limit expansion.

Ethereum Foundation

•

Cross-client research

Research & engineering infrastructure

The team collaborates with Ethereum Foundation researchers and execution-client teams on benchmarking infrastructure, execution readiness, and Ethereum scaling work.