Score: 0

Performance and Numerical Aspects of Decompositional Factorizations with FP64 Floating-Point Emulation in INT8

Published: September 28, 2025 | arXiv ID: 2509.23565v1

By: Piotr Luszczek , Vijay Gadepally , LaToya Anderson and more

Potential Business Impact:

Makes computers faster and use less power.

Business Areas:
DSP Hardware

Mixing precisions for performance has been an ongoing trend as the modern hardware accelerators started including new, and mostly lower-precision, data formats. The advantage of using them is a great potential of performance gain and energy savings. The disadvantage are the numerical issues not present in the standard-mandated floating-point formats. Split integer emulation of FP64 takes this to an extreme with the computation performed only by fixed-point tensor core units. We present the new issues the emulation faces for practical cases involving dense linear solver. We show extensive numerical tests indicating the effect of extended numerical range of matrix entries. We also scaled the input sizes to study the performance and numerical profiles on the NVIDIA Hopper GPUs.

Page Count
7 pages

Category
Mathematics:
Numerical Analysis (Math)