OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
ZK & Privacy

Sppark

Sppark is Supranational’s open-source CUDA/C++ template library for accelerating core operations used to generate SNARK and STARK zero-knowledge proofs.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Sppark — OpenViST project profile in ZK & Privacy.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Sppark when a custom zero-knowledge application needs low-level, GPU-accelerated proving primitives and the development environment can accommodate CUDA/C++ integration.

Who it is for

  • Developers building zero-knowledge proving systems
  • Teams integrating GPU acceleration into Rust, Go or C++ proof applications
  • Researchers and engineers benchmarking elliptic-curve and finite-field proof operations

Documented uses

  • Accelerate multi-scalar multiplication in proof-generation pipelines
  • Run number theoretic transform kernels for zero-knowledge workloads
  • Reuse finite-field, elliptic-curve, arithmetic-hash and related primitives
  • Integrate CUDA/C++ proving components into Rust or Go applications
  • Explore proof-of-concept benchmarks for supported operations
Before you choose

Requirements & limits

The repository is publicly accessible and licensed under Apache License 2.0. No account, pricing or service-availability requirement is documented in the supplied material.

  • Nvidia Volta-or-newer hardware is the primary GPU target.
  • Linux and Windows are the primary supported operating systems for the main GPU path.
  • AMD RDNA and CDNA support is limited.
  • Non-GPU portions can be used on ARM64 and macOS.
  • Rust and Go interfaces are shown, but host-language memory management, asynchronous behavior and data layout may require additional integration work.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up a Rust or Go integration starting point

Prepare an application to interface with Sppark’s accelerated zero-knowledge primitives.

Before you start

  • A Rust or Go application
  • A target environment supported by the repository, preferably x86_64 with an Nvidia Volta-or-newer GPU on Linux or Windows
  • The Sppark repository and its CUDA/C++ toolchain context

Steps

  1. Review the repository’s platform and language compatibility notes.

  2. Choose the relevant Sppark primitives, such as MSM, NTT, finite-field, elliptic-curve or hash components.

  3. Use the repository’s Rust or Go interface examples as the integration starting point.

  4. Adapt memory management, asynchronous operations and data layout to the host language where required.

Expected outcome

A language-integration starting point for incorporating Sppark primitives into a zero-knowledge application.

Guide limitations

  • The introductory integration and MSM tutorial sections are marked as to be determined.
  • The repository warns that target-language integration may require case-by-case changes.
  • Limited AMD support may affect available coverage and performance.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 October 2026
Original website title
supranational/sppark · Official homepage title · retained in its source language
How we review sourcesSuggest a correction