onnxruntime/cmake/external
Ted Themistokleous a561fde126
MIGraphX Execution Provider: Stream Synchronization (#12899)
**Description**: Changes to the MIGraphx execution provider code to
allow for stream synchronization on the gpu side

**Motivation and Context**
Performance boost by removing redundant host to device synchronizations 

The current implementation of the execution provider continuously calls
hipDeviceSynchronize() between computations which adds overhead and an
idle wait between the GPU's computations. This is noticeable during
device

This change leverages new functionality that's been added to MIGraphX to
allow for GPU side synchronization which avoids the need for
host->device waits.

To maintain backwards compatibility with older MIGraphX versions, the
compile time define MIGRAPHX_STREAM_SYNC has been added to the API to
allow for older version operate with newer builds of onnxruntime without
loss of functionality to the current feature set as of (08/09/22)

Co-authored-by: Ted Themistokleous <tthemist@amd.com>
2022-10-14 10:23:51 -07:00
..
cub@c3cceac115
cxxopts@3c73d91c0b
date@e7e1482087
dlpack@2775088798
eigen@d10b27fe37
emsdk@fc645b7626 MIGraphX Execution Provider: Stream Synchronization (#12899) 2022-10-14 10:23:51 -07:00
flatbuffers@6df40a2471
FP16@0a92994d72 XNNPACK EP (#11445) 2022-06-03 20:22:34 +10:00
googlebenchmark@7d0d9061d8
googletest@53495a2a7d
json@b5364faf9d Support per-test tolerances for ONNX tests (#11775) 2022-06-14 15:12:23 -07:00
libprotobuf-mutator@7a2ed51a6b
mimalloc@f412df7a2b
mp11@21cace4e57
nsync@436617053d
onnx@f7ee1ac60d MIGraphX Execution Provider: Stream Synchronization (#12899) 2022-10-14 10:23:51 -07:00
onnx-tensorrt@87c7a70688 [TensorRT EP] support TensorRT 8.4 (#11866) 2022-06-16 07:46:40 -07:00
onnxruntime-extensions@d4b2aff0c8
protobuf@0dab03ba7b MIGraphX Execution Provider: Stream Synchronization (#12899) 2022-10-14 10:23:51 -07:00
pthreadpool@1787867f61 XNNPACK EP (#11445) 2022-06-03 20:22:34 +10:00
pytorch_cpuinfo@5916273f79
re2@4244cd1cb4
SafeInt revert safeint version (#11500) 2022-05-12 11:24:43 -07:00
tensorboard@373eb09e4c
wil@e8c599bca6
XNNPACK@003c580e69 [xnnpack] basic QDQ operators support (#11912) 2022-08-11 10:12:51 +08:00
abseil-cpp.cmake Remove dynamic allocation for ThreadPool ParallelSection (#12429) 2022-08-04 09:46:16 -07:00
abseil-cpp.natvis Natvis adjustments to make debugging bearable (#13237) 2022-10-10 10:06:55 -07:00
composable_kernel.cmake Update ROCm CI to use HIP LANGUAGE (#13214) 2022-10-05 16:15:16 +08:00
dml.cmake Update DML 1.9.0 to 1.9.1 (#12966) 2022-09-15 10:54:22 -07:00
dnnl.cmake Upgrade the oneDNN ep to use oneDNNv2.7 (#13175) 2022-09-30 12:29:17 -07:00
eigen.cmake
extensions.cmake enable the extensions custom build for java and android (#11823) 2022-07-05 10:34:14 -07:00
FindNumPy.cmake
gsl.natvis Natvis adjustments to make debugging bearable (#13237) 2022-10-10 10:06:55 -07:00
ipp-crypto.cmake [TVM EP] Hot fix of build on Windows of TVM EP with ipp-crypto (#12381) 2022-07-31 14:36:54 +02:00
jemalloc.cmake
mimalloc.cmake
onnx_minimal.cmake Enable transpose optimizer in minimal extended build (#10349) 2022-01-31 09:41:04 -08:00
onnx_protobuf.natvis Natvis adjustments to make debugging bearable (#13237) 2022-10-10 10:06:55 -07:00
pybind11.cmake
pyxir.cmake
tvm.cmake Drop nuphar from java API (#13107) 2022-09-26 17:06:08 -07:00
xnnpack.cmake fix XNNPACK on WebAssembly SIMD (#13161) 2022-09-30 16:34:15 -07:00
zlib.cmake