The Information Machine
ConcludedRan 1 day2 sourcesupdated 21 Oct 2026

NVIDIA cuDSS Library Goes GA; Mistral Partnership Targets Industrial AI

The gist

GPU-accelerated sparse solvers can reduce simulation time and expand the number of designs teams can test. The Mistral-NVIDIA partnership targets open benchmarks for physics AI and agentic AI evaluations in industrial CAE and EDA workflows.

When stories like this move, I write about them in an email each morning.

One free email each morning, around 8 AM US Eastern.

The full picture

NVIDIA has released cuDSS, a CUDA-X library that accelerates sparse linear system solving by offloading the computation to GPUs. Sparse system solving is a repeated bottleneck in engineering and scientific simulations. Applied Materials reported up to 10x speedup in its Ginestra semiconductor simulation platform, ZEISS SMT reported a 4x speedup in an EUV lithography plasma simulation using COMSOL Multiphysics, and Dassault Systèmes reported 8-10x speedup in Abaqus electromagnetic solvers. cuDSS supports C, C++, Python, and Fortran interfaces and scales from workstations to large computing clusters. Separately, Mistral and NVIDIA announced a collaboration to bring AI to industrial engineering workflows, including faster simulation capabilities, open benchmarks and datasets for physics AI, and evaluations for agentic AI in computer-aided engineering and electronic design automation.

How it developed
21 October 2026

Mistral and NVIDIA announce collaboration on AI for industrial engineering, including physics AI benchmarks and agentic AI evaluations

Sources

Get the daily email

When stories like this move, I write about them in an email each morning.

One free email each morning, around 8 AM US Eastern.

See the latest edition · Privacy