We work on practical infrastructure for zero-knowledge proofs to achieve trustless computations at scale.
Distributed computing achieves horizontal scaling through massive parallelism across a network of nodes, and we show that it can be effectively applied to ZK proving to reduce both latency and cost. In this technical report, we will provide an overview of our distributed proving methodology and go over benchmarks for PSE’s EVM circuit (k=18) using our latest distributed halo2 prover.
Zero knowledge proofs (ZKPs) are remarkable cryptographic constructs with exciting applications in privacy and verifiability. However, the difficulty of writing circuits and high cost of proving greatly hinders its adoption. Many promising startups in the industry are tackling the first challenge by creating zkVMs with automatic circuit generation, and we at Giro Labs are working hard to solve the second one with our decentralized proving protocol.
There are fundamentally two kinds of trust needed in Web 3: “Programmed Trust” ensured by consensus and “Cryptographic Trust” ensured by ZK. In this article, we will describe how these two pieces play a role and how the scaling of each will unlock a new era ready for mass adoption.
In our previous post, we introduced the idea of using AWS EC2 spot instances for complex ZKP workloads. Specifically, we described in detail a method for computing MSMs. In this article, we will focus on another computationally intensive task: Number Theoretic Transform (NTT).
It is difficult to compute zero-knowledge proofs (ZKPs) on commercial hardware given the high memory requirement, and doing it on the cloud using memory-optimized instances can quickly rack up hefty bills. One way to lower the cloud computing cost is to use AWS EC2 spot instances.
Hello, world! We are Giro Labs, a pioneering distributed proving infrastructure that provides specialized ZK computing . In this article, we will take a deep dive into what we do and what we believe in.