Technology / WORLD

NVIDIA makes cuDSS generally available for engineering and scientific simulations

The CUDA-X library uses GPUs to solve large sparse systems in engineering and scientific applications.

Reading sizeSources ↓Share story ↗
Conceptual illustration of technology, processors and connected information.Illustration
Symbolic illustration

NVIDIA announced general availability of its cuDSS software library on 21 October. Part of CUDA-X, it provides GPU acceleration for large sparse linear systems repeatedly solved in engineering and scientific simulations. Such calculations are used to model designs ranging from chips and aircraft to industrial systems.

NVIDIA cited integrations in products from Applied Materials, COMSOL, Cadence, Dassault Systèmes, Siemens and Synopsys. The speedup figures in the announcement are company-reported measurements for specific software and workloads, not a general performance guarantee. Developers can integrate the library through C, C++, Python and Fortran interfaces.

Data

Timeline

NVIDIA makes cuDSS generally available for engineering and scientific simulations

  1. · Occurred

    NVIDIA makes cuDSS generally available for engineering and scientific simulations

    The CUDA-X library uses GPUs to solve large sparse systems in engineering and scientific applications.

Methodology

Based on NVIDIA’s 21 October product announcement; no independent performance result is inferred.

Updated:

Structured coverage

Subjects and places

Subject proposals

Taxonomy version: IPTC Media Topics 2026-Q2

Continue reading

Technology

Ember changes data placement when workloads exceed GPU memory

Researchers at the Technical University of Munich developed Ember, a memory manager that tracks which data should remain on a GPU when its memory fills. In the paper’s data-intensive workloads, it reported higher speeds than standard Unified Virtual Memory on three NVIDIA GPUs.

arXiv1 min read Data included