Ozaki Scheme
An algorithmic technique that approximates high-precision results through integer or low-precision floating-point matrix multiply-accumulate operations. It originates from the precision-decomposition approach proposed by the Japanese researcher Ozaki and colleagues, and is commonly used to emulate FP64 computation on low-precision hardware.
Timeline (last 90 days)
NVIDIA's upcoming Vera Rubin platform relies heavily on FP64 emulation capability based on the Ozaki scheme, recovering double-precision results by splitting operations into lower-precision operations.
The Ozaki scheme is an algorithmic technique that assembles high-precision results from integer or low-precision floating-point matrix multiply-add operations, originating from the precision decomposition approach proposed by the Japanese scholar Ohtsuka and others.
All Facts (2)
NVIDIA's upcoming Vera Rubin platform relies heavily on FP64 emulation capability based on the Ozaki scheme, recovering double-precision results by splitting operations into lower-precision operations.
50%UnverifiedThe Ozaki scheme is an algorithmic technique that assembles high-precision results from integer or low-precision floating-point matrix multiply-add operations, originating from the precision decomposition approach proposed by the Japanese scholar Ohtsuka and others.
50%