Risc Zero announced today the launch of zkVM 1.0, a high-performing and production-ready zero-knowledge proof virtual machine.
zkVM 1.0 makes zero-knowledge proofs scalable, accessible, and cost-effective for any chain.
Risc Zero
Beating Moore’s Law with zkVM 1.0
zkVM 1.0 allows developers to prove any Rust program and verify them on any chain with the goal of increasing inter-chain compatibility to unleash a new wave of complex dApps across the blockchain ecosystem.
Risc Zero’s zkVM 1.0 offers high performance, surpassing competitors in general-purpose computations. Continuous improvement ensures that Risc Zero will maintain its performance lead in zkVM. Risc Zero has shared comprehensive benchmarks to demonstrate its cost and speed superiority across various workloads.
Performance of Risc Zero’s zkVM 1.0
Recent benchmarks highlight Risc Zero’s advantage over Succinct’s SP1. Using a variety of hardware configurations, ranging from consumer devices to cloud instances, Risc Zero demonstrated superior performance under fair conditions with all performance features enabled for both systems.
The tests were conducted on multiple cloud instances and consumer devices, with various hardware configurations including RISC-V execution, random access memory, and SHA2 accelerators. Comparisons were made using recommended performance indicators to ensure fairness.
Key Results from Risc Zero’s zkVM 1.0 Benchmarks
The results show that Risc Zero’s zkVM outperforms Succinct’s SP1 in terms of cost and speed, whether on consumer devices or in the cloud. On a g6.xlarge cloud instance, Risc Zero is both faster and cheaper than an r7i.16xlarge instance used by SP1. For certain workloads, Risc Zero is up to 60 times cheaper than SP1.
However, in specific cases such as crypto operations, Succinct’s SP1 demonstrates faster performance due to the use of accelerators. Risc Zero claims to be actively working on integrating similar accelerators to enhance performance in these scenarios.
Risc Zero has made its benchmarking scripts and data available to ensure transparency and reproducibility of results. Users can refer to the guide and raw data to reproduce the results.
