Lollipop: SVM Rollups on Solana
Irvin Steve Cardenas, Yugart Song
TL;DR
The paper defines a formal specification for SVM rollups on Solana L1, addressing scalable execution, data availability, and fraud-proof mechanisms. It leverages Sparse Merkle Trees for state commitments and KZG polynomial commitments for data, coupled with an interactive fraud-proof protocol and transaction replay to ensure correctness. Two node roles—execution nodes and validator nodes—together enable secure rollup operation with a data availability committee (DAC) and data auditing/sampling to maintain data integrity. It also outlines practical considerations, including hardware requirements, cross-L1 compatibility via zk-proofs, and future directions such as decentralized sequencers and extending the approach to Bitcoin L2. Collectively, the work aims to enable trust-minimized, scalable SVM rollups on Solana while outlining pathways for broader applicability across other L1 ecosystems.
Abstract
We present a formal specification for the implementation of Solana virtual machine (SVM) rollups deployed on top of the Solana Layer 1 (L1) blockchain. We further discuss our motivation, implementation, design decisions, limitations, and preliminary results. Overall, this paper is intended to serve as an initial introduction to building such system(s) on top of the Solana L1 blockchain, but does not represent an absolute. Lastly, we comment discuss on extensions of this specification to support SVM rollups on other well-established L1 blockchains systems such as Ethereum.