educative
In this blog post, we explore an optimization for Barrett reduction, a popular method for modular arithmetic. We discovered that the error bound can be tighter than traditionally thought, meaning you often only need a single subtraction instead of two. This tweak can speed up cryptographic operations like those in the RustCrypto library by 14% when applied to NIST curves, which is a game changer for both encryption and performance. Dive into the details to see how this simple change can lead to significant improvements in real-world applications!
Suneal Gong
May 01, 2025
10 min
Read →
Variants of KZG · Part 1
In this blog post, we dive into the world of polynomial commitment schemes (PCS), which are crucial for constructing most practical SNARKs. We cover the basics of how PCS works, focusing on KZG10, known for its efficiency in proof size and verification time. You'll learn about the essential properties of binding and hiding and explore technical concepts like homomorphism, batching, and unconditionally hiding. We break down various methods to achieve these features, offering insight into how PCS maintains the security and privacy of polynomials in cryptographic systems. Get ready to understand these powerful concepts and their applications in modern cryptography!
Varun Thakore
April 28, 2025
17 min
educative · tools
Hey there! We're diving into how leveraging WebGPU can revolutionize client-side proving for privacy-preserving zero-knowledge apps. By tackling the usual time and memory bottlenecks with innovative techniques like using GPUs, we're achieving significant performance gains. We'll share our experiences using WebGPU with various frameworks, offering insights into optimizing operations and managing data transfer efficiently. Plus, learn about the challenges and exciting potential of integrating WebGPU into zero-knowledge frameworks. Join us to explore the future of privacy tech!
Jaehun Kim, Jason Park
April 25, 2025
20 min
