Optimum’s mump2p Doubles Ethereum Block Delivery Speed

Optimum’s mump2p goes live on Ethereum mainnet

Ethereum validators have started using Optimum’s mump2p peer-to-peer networking protocol on mainnet, and early reports suggest materially faster block propagation across participating nodes. The upgrade does not change Ethereum’s consensus rules; instead, it provides an additional networking layer that runs alongside existing consensus clients to accelerate delivery of blocks and attestations using Random Linear Network Coding (RLNC).

How mump2p works: RLNC and the gateway approach

Optimum’s mump2p uses Random Linear Network Coding (RLNC) to reduce redundant data transfers and speed up block delivery across the Ethereum peer-to-peer network. Rather than forwarding identical copies of the same block through GossipSub, mump2p transmits coded combinations of block fragments; receiving nodes can reconstruct the original block once they obtain enough independent coded packets.

Gateway integration and client compatibility

The protocol operates through a gateway that sits next to a validator’s consensus client. This gateway bridges the standard Ethereum network and Optimum’s separate communication overlay, allowing validators to keep using Prysm, Lighthouse, Teku, Nimbus, Lodestar and other clients without altering consensus software or protocol rules. Because mump2p is additive, validators continue to receive messages over GossipSub while also benefiting from the coded broadcast path.

Measured performance: faster block propagation and reduced variance

Optimum’s initial mainnet launch on October 20 reported block propagation improvements ranging from roughly two to four times faster than the GossipSub baseline. Company data across early measurements indicated a 7.88-fold reduction in arrival-time variance and showed mump2p delivering blocks earlier than the standard route in nine out of ten samples.

A broad analysis covering 50,292 Ethereum slots found an average delivery advantage of about 335 milliseconds, roughly a 3.3x improvement in that test. Lido’s testnet evaluation also reported earlier block arrivals — approximately 407 milliseconds sooner across a September 17 to October 15 window covering 180,327 slots — though Lido cautioned that testnet runs do not directly translate to mainnet performance figures.

Independent testing and mixed results

Not all independent analyses have produced uniformly optimistic results. StereumLabs’ assessment across 36 mainnet nodes found earlier blocks on Lighthouse, Prysm and Grandine, little change for Teku, and slower arrivals for Nimbus. Their sample did not show statistically significant overall improvements.

Optimum responded by noting its global acceleration layer may have been active during some tests, which could have allowed untreated nodes to indirectly benefit and masked the protocol’s differential impact. Testing across diverse network conditions, client configurations, and geographies remains important to quantify end-to-end benefits accurately.

Adoption: who’s running mump2p and stake participation

At launch, Optimum reported 16 node operators connecting roughly 22,000 validator keys and more than 1.4 million staked ETH — a figure Optimum described as Total Value Boosted (TVB), worth about $3.4 billion at the time. By November 3, adoption had grown to cover an estimated 7.8% of all staked ETH, up from just over 3% at the October debut.

Participating operators included major infrastructure players such as Blockdaemon, Everstake, P2P.org, Stakefish, Luganodes, InfStones, Validation Cloud, and provider-specific integrations — for example, restaking provider ether.fi required its node operators to use mump2p. Stakefish confirmed its validator stack integrated Optimum’s gateway, while the service remained free for validator block propagation, with operators onboarding through Optimum’s gateway process.

Why this matters for validators and the Ethereum network

Faster and more reliable block propagation reduces the chance of stale or orphaned blocks and can improve proposer and attester coordination across the network. For large-scale validator operators, lower variance in arrival times and consistent earlier delivery can translate into higher performance, fewer missed attestations, and potentially better rewards capture.

Because mump2p is non-invasive to consensus logic, validators can trial the system without risking protocol incompatibility. This makes the gateway model attractive for operators who want to test improvements to P2P networking, block propagation, and overall validator throughput while keeping their existing consensus clients intact.

Tradeoffs and open questions

Network-coded broadcasting introduces computational and bandwidth tradeoffs — encoding and decoding coded packets require CPU and memory resources, and the overlay must manage peer selection and redundancy. Moreover, results can vary by client implementation and geographic distribution of nodes. Continued cross-client testing, long-run mainnet measurements, and third-party audits will be important to validate claims at scale.

Alignment with Ethereum’s roadmap: erasure-coded networking and post-quantum research

Ethereum co-founder Vitalik Buterin has publicly highlighted erasure-coded P2P networking as a potential improvement for delivering large data objects and for long-term security goals. In his posts about Ethereum’s roadmap, Buterin described erasure-coded networking among several proposals that could support faster consensus and even post-quantum signature aggregation strategies aimed at shorter finality windows.

These broader roadmap items — including EIP-8411 experiments on segmented block propagation and research into shorter slots, few-slot finality, and quantum-resistant signatures — remain separate from Optimum’s mump2p deployment. Implementing such consensus-level changes would require extensive protocol development and community consensus.

Bottom line

Optimum’s mump2p introduces network coding-based block propagation to Ethereum validators via a gateway that coexists with existing consensus clients. Early results show meaningful latency reductions and lower variance in block arrival times for participating validators, though independent tests have reported mixed outcomes. As adoption grows and more third-party evaluations become available, validators and node operators should watch how mump2p interacts with different client stacks and network topologies. If consistent at scale, network-coded P2P layers like mump2p could become an important tool for improving validator performance and overall network robustness as Ethereum continues to evolve toward faster finality and enhanced data availability.