Skip to content

Latest commit

 

History

History
64 lines (36 loc) · 862 Bytes

File metadata and controls

64 lines (36 loc) · 862 Bytes
title description author ms.author ms.date ms.service ms.subservice ms.topic keywords ms.devlang monikerRange
gpu_math: gpu_math
NVidia CUDA implementation.
VanMSFT
vanto
07/15/2019
sql
machine-learning-services
reference
neural network
math
gpu
python
>=sql-server-2017||>=sql-server-linux-ver15

microsoftml.gpu_math: Acceleration with NVidia CUDA

Usage

microsoftml.gpu_math(gpu_id: numbers.Real = -1,
    cu_dnn: bool = False, cu_dnn_algo: str = 'ImplicitPrecompGemm')

Description

NVidia CUDA implementation.

Arguments

gpu_id

GPU device id (settings).

cu_dnn

Use cuDNN on GPU (settings).

cu_dnn_algo

cuDNN optimization options (settings).

See also

avx_math, clr_math, mkl_math, sse_math