New research highlights a fundamental bottleneck in most Ethereum rollups: nearly all operate on a single-threaded EVM, forcing every transaction into one global queue. This architecture prevents true parallel execution, meaning busy applications can slow down the entire network, even on Layer 2.
Research Points To Core Constraint In Ethereum Rollups
The analysis shows that single-threaded execution severely restricts throughput. With every smart contract competing in a linear sequence, rollups cannot execute multiple independent operations simultaneously. As demand spikes, so do fees, because all transactions share a single fee market.
This limitation has become more visible as L2 activity grows, raising questions about whether the current EVM-based design can scale to support mainstream adoption.
RESEARCH: Most Ethereum rollups use a single-threaded EVM where all transactions compete in one global queue, making parallel execution impossible.
Eclipse brings Solana’s Virtual Machine to Ethereum, creating separate lanes so busy apps don’t slow the whole network.… pic.twitter.com/g4I5BxafjS
— Cointelegraph (@Cointelegraph) December 2, 2025
Eclipse Introduces Solana’s SVM To Ethereum For Parallel Execution
Eclipse proposes a different model by bringing Solana’s Virtual Machine (SVM), and its Sealevel parallel runtime, directly to Ethereum settlement. The chart at the bottom highlights the contrast:
- •The single-threaded EVM processes contracts linearly.
- •SVM’s Sealevel runs multiple contracts at once, enabling true parallel execution.
Eclipse accomplishes this by isolating workloads into separate lanes, meaning high-traffic applications no longer degrade performance for the rest of the network. The data indicates that this structure offers a way for Ethereum L2s to scale horizontally rather than vertically.
Localized Fees Prevent Network-Wide Cost Spikes
Traditional rollups share a unified fee market, so one application’s surge can increase gas costs across the entire network. Eclipse breaks this pattern by using localized fees, allowing each lane to maintain its own cost structure.
As a result, isolated activity, such as a trading app experiencing sudden demand, does not raise execution costs for unrelated applications running elsewhere in the system.
Faster Settlement With ZK Fraud Proofs
Another notable shift is Eclipse’s use of zero-knowledge fraud proofs, which replace the multi-round interactive disputes common in optimistic rollups. This significantly reduces settlement time and lowers the operational overhead associated with fraud detection.
Pairing Solana-style parallelism with Ethereum settlement provides a hybrid model: high throughput with strong security guarantees.
The Central Question: Can Ethereum L2s Achieve True Parallelism?
The research raises a broader discussion across the ecosystem: Can Ethereum rollups embrace meaningful parallel execution without compromising determinism and settlement guarantees?
Eclipse presents one possible answer by combining Solana’s proven parallel runtime with Ethereum’s verification model. If successful, it could mark a shift in how L2s scale, moving away from single-thread constraints and toward architectures capable of supporting high-volume, multi-application environments.
As demand on Ethereum continues to rise, the debate over parallelism versus EVM compatibility is set to intensify, shaping the next generation of rollup design.