FP64 Performance and the Limits of Ozaki Emulation on HPC Systems
cs.DC, cs.NA, math.NA
Submitted: 2026-11-09
Updated: 2026-11-09
Code: https://github.com/kswirydo/ozaki
Terminology
Sources
- Guaranteed DGEMM Accuracy While Using Reduced Precision Tensor Cores Through Extensions of the Ozaki Scheme
- EmuGEMM: Fused Tensor Core Kernels for Precision Emulation in Matrix Multiplication
- Analysis of Floating-Point Matrix Multiplication Computed via Integer Arithmetic
- Error Analysis of Matrix Multiplication Emulation Using Ozaki-II Scheme
- FP8 is All You Need (Part 1): Debunking Hardware FP64 as the HPC Holy Grail (Sep 3rd version)
- Improved Scaling for Fast Mode of Ozaki Scheme II
- Double-Precision Matrix Multiplication Emulation via Ozaki-II Scheme with FP8 Quantization
Related papers
- Programmable Probabilistic Computer with 1,000,000 p-bits
- BatchGen: An Architecture for Scalable and Efficient Batch Inference
- Seamless Reconfiguration for DAG-BFT (Technical Report)
- ELDR: Expert-Locality-Aware Decode Routing for PD-Disaggregated MoE Serving
- iScheduler: Reinforcement Learning-Driven Continual Optimization for Large-Scale Resource Investment Problems
- TIRx: A Unified Tile Primitive ML Compiler Across Heterogeneous Backends