MEV on Ethereum vs Solana: Mempool Architecture and Opportunity Differences

MEV extraction on Ethereum and Solana are not just different in degree — they are structurally different games. The mempool architecture, block production pipeline, and competitive dynamics on each chain produce distinct opportunity profiles and infrastructure requirements. If you are deciding where to invest your searcher stack, the comparison is worth working through in detail.

Ethereum: The Public Mempool and PBS Pipeline

Ethereum's public mempool is the starting point for most MEV discussion. Transactions broadcast to the network sit in a publicly visible pending state, giving any observer a window into pending state changes before they are included in a block. This visibility is the root of most sandwich attacks, liquidation races, and DEX arbitrage that defined early MEV.

Proposer-Builder Separation (PBS) via MEV-Boost restructured how that opportunity is captured. Today, roughly 90% of Ethereum blocks are built by a handful of specialized builders — Beaverbuild, Titan, rsync — who receive bundles from searchers, assemble optimal blocks, and route them to validators via relays. The relay acts as a trusted intermediary: the validator commits to using the builder's block in exchange for the bid, without seeing the full transaction content until the slot is won.

This creates a clear two-tier market. Searchers compete to send profitable bundles to builders; builders compete to offer the highest bid to validators. Your edge as a searcher on Ethereum comes down to simulation speed, bundle ordering logic, and relationships with builders. Latency to a co-located simulation node matters, but so does understanding which builder is likely to include your bundle given their current block composition.

Key Ethereum MEV characteristics:

• Pending transactions visible in the public mempool (~12s block time gives significant search window)
• Bundle submission via eth_sendBundle to builder endpoints (Flashbots, BloXroute, private relays)
• Atomic bundle execution guarantees — your bundle lands entirely or not at all
• Reverting transactions do not cost gas if structured correctly inside a bundle
• Dominant strategies: sandwich attacks, cross-DEX arbitrage, liquidations, just-in-time liquidity

Solana: Gulf Stream and the Absence of a Traditional Mempool

Solana does not have a mempool in the Ethereum sense. Gulf Stream is the protocol mechanism that forwards transactions directly to the current and upcoming leaders — the validators scheduled to produce blocks in the next several slots. Transactions do not pool in a publicly visible pending state. They are routed forward and either included in a block or dropped.

This has a fundamental consequence: there is no canonical pending transaction queue to monitor and front-run. The pre-confirmation information advantage that powers most Ethereum sandwich attacks simply does not exist in the same form. Solana's 400ms slot time compounds this — the search window before a transaction confirms is measured in milliseconds, not seconds.

What replaced the mempool: Jito's tip auction. Jito Labs deployed a modified validator client that runs an off-chain block engine. Searchers submit bundles to Jito's block engine with a tip denominated in SOL. The block engine auctions bundle ordering within a slot, and winning bundles are included atomically at the top of the block. As of 2025, Jito processes the majority of MEV-sensitive transactions on Solana, and tip levels for competitive opportunities regularly exceed $50-200 per successful capture.

Latency Profile and Infrastructure Differences

On Ethereum, your simulation infrastructure can sit anywhere with reasonable connectivity to builder endpoints. A 10-50ms round trip to a major builder is workable. The 12-second block time means the competitive window is wide enough that raw simulation throughput matters more than microsecond latency.

On Solana, the infrastructure picture is different. With 400ms slots and a direct-to-leader forwarding model, co-location with Jito's block engine and proximity to the current leader's staked validator nodes is meaningful. Searchers running serious Solana MEV infrastructure maintain nodes in multiple geographic regions and dynamically route based on the leader schedule. Transaction fees are low enough that volume of attempts is also a viable strategy for some opportunity types.

Opportunity Type Comparison

The structural differences produce different dominant opportunity categories:

Opportunity	Ethereum	Solana
Sandwich attacks	Common, mempool-visible	Rare, requires Jito bundle timing
DEX arbitrage	Competitive, bundle-based	Very competitive, latency-driven
Liquidations	Large protocols, gas auctions	Fast, tip auction determines ordering
JIT liquidity	Active, UniV3 focused	Emerging on Orca/Meteora
Cross-chain arb	Bridging latency limits	Faster finality advantages

Solana arbitrage in practice is dominated by cyclic routes across Raydium, Orca, Jupiter-aggregated pools, and the newer concentrated liquidity venues. The opportunity arises from price discrepancies that open and close within a single slot. Speed of detection and bundle submission latency to Jito's block engine are the primary differentiators.

Where Searchers Should Actually Allocate

Ethereum remains the higher absolute-value chain for MEV — total extractable value per day is larger, and the bundle infrastructure is more mature. But it is also more competitive and capital-intensive, with sophisticated builders and well-funded searcher teams dominating the top of the fee ladder.

Solana's MEV market is younger and the tooling less standardized, which creates opportunities for searchers willing to work closer to the metal. If you are comfortable with Rust, understand the Solana account model, and can navigate Jito's bundle API, the competitive density in specific opportunity niches is lower than the equivalent Ethereum strategy.

The practical answer for most teams: Ethereum for large-capital liquidation and arbitrage where established infrastructure de-risks execution; Solana for high-frequency, lower-capital opportunities where latency advantage and novel pool mechanics create edge. Many teams we work with at our trading bot practice run both simultaneously, treating the chains as complementary rather than competing allocation decisions.

---

If you are evaluating where to deploy a MEV strategy or want an honest assessment of the infrastructure requirements on either chain, reach out directly — we can scope a build based on your capital, risk tolerance, and target opportunity set.