kroma-network/tachyon
Tachyon
Tachyon is a suite of hardware-accelerated engines and modular infrastructure for generating succinct non-interactive arguments of knowledge and polynomial commitments. It provides a GPU-accelerated zero-knowledge proof backend and a prover system designed to construct proofs from high-level circuit descriptions.
The project distinguishes itself through a GPU cryptographic accelerator that offloads elliptic curve operations, field arithmetic, and cryptographic hashes to graphics hardware. This infrastructure includes a specialized polynomial commitment engine and modular components that allow for interoperability across diverse proof systems and frontends.
The system covers broad capability areas including finite field arithmetic, fast Fourier transforms, and cryptographic commitment schemes. It also includes low-overhead execution trace profiling to identify system bottlenecks during the proof generation process.
Features
- Provers - Provides a complete system for constructing succinct non-interactive arguments of knowledge from high-level circuit descriptions.
- GPU-Accelerated Generation - Uses GPU hardware to significantly reduce the time and computational resources required to generate zero-knowledge proofs.
- Polynomial Commitments - Implements cryptographic methods to bind data to polynomials, allowing efficient verification without revealing the full dataset.
- Cryptographic Polynomial Computation - Performs fast Fourier transforms and polynomial commitments to optimize the processing of large datasets for ZK circuits.
- Fourier Transforms - Processes large data arrays using specialized algorithms to optimize the multiplication of complex polynomials.
- Radix-Based Transforms - Processes large polynomial datasets using Radix2 and MixedRadix algorithms to optimize multiplication and convolution speed.
- Finite Field Arithmetic - Executes mathematical operations over multiple field types using CPU or GPU resources for optimized computation.
- Finite Field Arithmetic - Implements high-throughput mathematical operations over finite fields by offloading computations to graphics processing units.
- Cryptographic Operation Accelerators - Provides a library for performing elliptic curve operations, field arithmetic, and cryptographic hashes on graphics hardware.
- Elliptic Curve Cryptography - Provides a hardware-accelerated system for computing points and scalars on standard curves.
- Hardware-Accelerated Operations - Computes point and scalar multiplications on cryptographic curves using specialized hardware instructions to speed up proof generation.
- Zero-Knowledge Proof Frameworks - Provides modular cryptographic infrastructure that handles field arithmetic and elliptic curve operations for proof systems.
- Modular Proof Infrastructure - Provides a set of interoperable components for processing proof frontends and commitment schemes across diverse proof systems.
- GPU-Accelerated Backends - Ships a hardware-accelerated backend for generating zero-knowledge proofs using GPU resources.
- Modular Frontend Integrations - Interfaces with various zero-knowledge circuit descriptions to translate high-level proofs into executable hardware-accelerated operations.
- Commitment Schemes - Ships a specialized engine for generating cryptographic polynomial commitments to secure and verify data integrity.
- SNARK Provers - Constructs succinct non-interactive arguments of knowledge by processing high-level circuit descriptions through accelerated backends.
- Zero-Knowledge Proof Generation - Creates succinct non-interactive arguments of knowledge by processing high-level circuit descriptions through accelerated backends.
- GPU Accelerators - Uses GPU backends to significantly reduce the time and computational resources required for generating zero-knowledge proofs.
- Zero-Knowledge Hash Functions - Provides collision-resistant hash functions specifically optimized for use within zero-knowledge proof circuits.
- Frontend Integrations - Provides modular components that interoperate with diverse proof systems to translate high-level circuits into accelerated operations.
- Execution Tracing - Captures detailed performance data during proof generation to identify and resolve system bottlenecks.
- ZK Proof Performance Profiling - Analyzes execution traces and system bottlenecks to optimize the speed of cryptographic proof generation.
