Scaling blockchain performance through parallel smart contract execution has gained significant attention, as traditional methods are limited by the performance of a single virtual machine (VM), even in multi-chain or Layer-2 systems. Parallel VMs present a promising solution by enabling concurrent execution of transactions within a single smart contract, utilizing multiple CPU cores. However, existing parallel mechanisms are constrained by Ethereum’s sequential, shared-everything model, which does not fully exploit the parallelizable aspects of smart contracts. This leads to frequent rollbacks with optimistic parallelization and substantial overhead in pessimistic approaches due to state dependency analysis and locking.

In this paper, we introduce Crystality, a programming model for smart contracts on parallel Ethereum Virtual Machines (EVMs) that enables developers to express the inherent parallelism within a smart contract. Crystality introduces Programmable Contract Scopes to partition contract states into non-overlapping, parallelizable segments and decompose a smart contract function into finer-grained components. Crystality also features Asynchronous Functional Relay to manage execution flow across EVMs. These innovations simplify the expression of parallelism and allow the parallel VM system to perform asynchronous execution when operations on a contract state are commutative.

Crystality extends Solidity with directives and transpiles Crystality code into standard Solidity code, ensuring EVM compatibility. The Crystality system includes two execution modes: an asynchronous mode for transactions involving commutative operations and an optimistic-based fallback mode to maintain block-defined transaction order. Our experiments demonstrated Crystality’s superior performance compared to Ethereum, Aptos, and Sui on a 64-core machine.