Blog

Releases, technical deep dives, and posts from ecosystem partners.

GPU-Accelerated Big Integer Arithmetic and NTT in cuPQC 0.6

The 0.6 math layer: multi-precision arithmetic up to 4096 bits and transforms up to 2²⁴ points, with benchmarks and the configuration choices behind them.

Read post →

Published August 28, 2026

Release v0.6

Topics

BigIntNTT

Ecosystem Aug 29, 2026

Accelerating Encrypted AI with Lattica

Integrating cuPQC NTT kernels into the Lattica Platform and HEAL runtime, with encrypted MNIST inference reaching 7,849 inferences per second.

LatticaNTTFHE

Read post →

Ecosystem Aug 29, 2026

FHERMA Challenges for GPU-Accelerated FHE

Two new FHERMA challenge tracks invite GPU developers to build reusable FHE primitives with cuPQC: high-degree polynomial arithmetic and CKKS key switching.

Fair MathBigIntFHE

Read post →

Technical Blog Apr 9, 2026

Merkle Trees with Poseidon 2

GPU-accelerated Poseidon 2 Merkle tree generation in cuPQC-Hash with Single-Block and Multi-Block modes, showing speedups up to 169× vs CPU.

Benchmarkv0.6

Read post →

Technical Blog Apr 7, 2026

Hash Performance (SHA2, SHA3, SHAKE)

CPU vs GPU throughput for SHA2, SHA3, and SHAKE with cuPQC-Hash on an NVIDIA RTX PRO 6000 versus an AMD EPYC 9124.

Benchmarkv0.4.1

Read post →

On the NVIDIA Developer Blog

v0.4 Improve Data Integrity and Security with Accelerated Hash Functions and Merkle Trees cuPQC v0.4 · August 21, 2025 NVIDIA Blog ↗ v0.2 Introducing NVIDIA cuPQC for GPU-Accelerated Post-Quantum Cryptography cuPQC v0.2 · December 3, 2024 NVIDIA Blog ↗