Execution — Decoupled SVM
The SVM’s parallel processing capability allows SOON Chains to handle a high volume of transactions, significantly improving throughput and scalability. However, in order to better fit the rollup use case, it was necessary for us to come up with something other than a Forked SVM which would have made it impossible to have a proper derivation layer and it would be incredibly difficult to implement fraud proofs. Our solution to this problem was to develop the Decoupled SVM framework which underpins the SOON Stack. Specifically, we decoupled the Transaction Processing Unit (TPU) in order to remove the original Solana consensus and allow for the TPU to be controlled by the rollup node, for derivation purposes. We are also able to remove vote transactions (which are not necessary for L2s), which leads to reduced data availability usage. In short, a Decoupled SVM is essential for derivation, brings greater network security and reduces DA waste.
Optimized Parallel Execution: SVM’s ability to execute non-overlapping transactions in parallel significantly enhances network throughput and reduces latency, allowing the system to handle a large number of transactions efficiently. This parallel execution model is particularly beneficial for high-frequency trading applications, decentralized exchanges, and other use cases requiring fast transaction finality.
Local Fee Markets: Solana’s local fee markets prevent congestion by isolating fee spikes to individual applications, ensuring a stable and predictable fee structure that benefits all users and maintains network stability. This model allows for differentiated fee structures tailored to the specific needs of various applications, providing flexibility and preventing fee market distortions.
Decoupling the Transaction Processing Unit (TPU): To better fit the rollup use case, SOON developed the Decoupled SVM framework. The key innovation here is the decoupling of the Transaction Processing Unit (TPU) from the original Solana consensus, allowing the TPU to be controlled by the rollup node for derivation purposes. This decoupling enables SOON to overcome the limitations of a traditional Forked SVM, such as the inability to implement a proper derivation layer and the challenges in creating fraud proofs.
Enhanced Network Security: By separating execution from consensus, the Decoupled SVM not only increases TPS but also strengthens network security. With the TPU controlled by the rollup node, the Decoupled SVM ensures the accuracy and reliability of the derivation process.
Censorship-resistant: By decoupling the SVM from Solana’s consensus mechanisms, the derivation pipeline gains the ability to control block production on L2 independently, ensuring that the L2 state is accurately derived from the L1 state without being restricted by Solana’s original consensus protocols. This feature enables users to enforce withdrawals, thereby providing censorship resistance and ensuring that funds can be accessed even in the face of potential censorship.
Last updated