NVIDIA cuBLAS Adds Ozaki-II and 2M for Faster FP64 Matrix Math
NVIDIA Technical Blog says CUDA Toolkit 13.4.2 introduces the Ozaki-II and 2M algorithms in cuBLAS to accelerate double-precision real and complex matrix multiplication on integer Tensor Cores. Ozaki-II requires fewer integer multiplications than Ozaki-I, while 2M reduces a complex multiplication from four real multiplications to two. NVIDIA reports that cuBLAS can apply the acceleration automatically without application-code changes, with benchmark results for scientific-computing applications.
The algorithms map FP64 work onto faster integer Tensor Cores while retaining double-precision accuracy, potentially reducing matrix-computation time in HPC and AI workloads without requiring application changes.
Many scientific simulations, from molecular chemistry to materials science, spend the majority of their execution time computing high-precision dense general matrix multiplications (GEMM). These operations are called DGEMM for real 64-bit floating-point numbers and ZGEMM for complex 64-bit floating-point numbers. cuBLAS is widely used by HPC and AI applications to accelerate GEMM and other Basic Linear Algebra Subprograms (BLAS) on NVIDIA GPUs. Rapid advances in lower-precision hardware have motivated exploration of a new class of algorithms that deliver full double-precision results on the faster integer Tensor Cores.
With CUDA Toolkit 13.4.2, cuBLAS no longer has to choose between accuracy and speed. cuBLAS breaks each double-precision number into integer pieces, multiplies them on the GPU’s Tensor Cores, and reassembles the exact result. The answers keep full double-precision accuracy, but the work runs on faster hardware.
Two new algorithms make this more efficient: Ozaki-II needs fewer integer multiplications than the earlier Ozaki-I approach, and the 2M method cuts a complex matrix multiplication from four real multiplications to two. This builds on the Automatic Dynamic Precision (ADP) framework introduced in Unlocking Tensor Core Performance with Floating Point Emulation in cuBLAS, which delivers higher double-precision (FP64) matrix-multiplication performance by mapping FP64 arithmetic onto integer Tensor Cores while preserving native FP64 accuracy.
#NVIDIA#cuBLAS#CUDA Toolkit 13.4.2#Ozaki-II#2M#Automatic Dynamic Precision#NVIDIA VR NVL72#NVIDIA GB200#NVIDIA RTX PRO 6000#BerkeleyGW