projects / platform / upstream / tensorflow.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 2c105ac) | patch
[TF:XLA] Add INTEL MKL_DNN Conv2d method to XLA/CPU backend
authorTony Wang <tonywy@google.com>
Thu, 26 Apr 2018 20:30:15 +0000 (13:30 -0700)
committerTensorFlower Gardener <gardener@tensorflow.org>
Thu, 26 Apr 2018 20:33:48 +0000 (13:33 -0700)
commitb6189a23a5f6afa59ced097d7844d58c7fd24901
treef703a5450dca1a10bf41b95718f9b63053bbef98tree | snapshot
parent2c105ace934edce193669b55b13b64283caa24d7commit | diff
[TF:XLA] Add INTEL MKL_DNN Conv2d method to XLA/CPU backend

The INTEL MKL_DNN provides 32-bit Conv2d method. With INTEL_MKL flag set,
XLA backend emits runtime call to MKL_DNN Conv2d instead of Eigen.

PiperOrigin-RevId: 194445212
tensorflow/compiler/xla/service/cpu/BUILD diff | blob | history
tensorflow/compiler/xla/service/cpu/cpu_runtime.cc diff | blob | history
tensorflow/compiler/xla/service/cpu/cpu_runtime.h diff | blob | history
tensorflow/compiler/xla/service/cpu/ir_emitter.cc diff | blob | history
tensorflow/compiler/xla/service/cpu/runtime_conv2d_mkl.cc [new file with mode: 0644] blob
tensorflow/compiler/xla/service/cpu/runtime_conv2d_mkl.h [new file with mode: 0644] blob
tensorflow/compiler/xla/service/cpu/simple_orc_jit.cc diff | blob | history
Domain: Machine Learning / ML Framework;