Skip to content

Commit

Permalink
[HIPIFY][DNN] cuDNN 9.3.0 is the latest supported version
Browse files Browse the repository at this point in the history
  • Loading branch information
emankov committed Aug 8, 2024
1 parent eb55d11 commit 67c2626
Show file tree
Hide file tree
Showing 7 changed files with 18 additions and 16 deletions.
22 changes: 11 additions & 11 deletions docs/hipify-clang.rst
Original file line number Diff line number Diff line change
Expand Up @@ -545,7 +545,7 @@ LLVM >= 10.0.0

.. code-block:: shell
-DCUDA_DNN_ROOT_DIR=D:/CUDA/cuDNN/9.2.1
-DCUDA_DNN_ROOT_DIR=D:/CUDA/cuDNN/9.3.0
5. [Optional] Install `CUB 1.9.8 <https://github.com/NVIDIA/cub/releases/tag/1.9.8>`_ for ``CUDA < 11.0`` only;
for ``CUDA >= 11.0``, the CUB shipped with CUDA will be used for testing.
Expand Down Expand Up @@ -637,8 +637,8 @@ On Linux, the following configurations are tested:

* Ubuntu 14: LLVM 4.0.0 - 7.1.0, CUDA 7.0 - 9.0, cuDNN 5.0.5 - 7.6.5
* Ubuntu 16-19: LLVM 8.0.0 - 14.0.6, CUDA 7.0 - 10.2, cuDNN 5.1.10 - 8.0.5
* Ubuntu 20-21: LLVM 9.0.0 - 18.1.8, CUDA 7.0 - 12.3.2, cuDNN 5.1.10 - 9.2.1
* Ubuntu 22-23: LLVM 13.0.0 - 18.1.8, CUDA 7.0 - 12.3.2, cuDNN 8.0.5 - 9.2.1
* Ubuntu 20-21: LLVM 9.0.0 - 18.1.8, CUDA 7.0 - 12.3.2, cuDNN 5.1.10 - 9.3.0
* Ubuntu 22-23: LLVM 13.0.0 - 18.1.8, CUDA 7.0 - 12.3.2, cuDNN 8.0.5 - 9.3.0

Minimum build system requirements for the above configurations:

Expand All @@ -658,7 +658,7 @@ Here's how to build ``hipify-clang`` with testing support on ``Ubuntu 23.10.01``
-DCMAKE_INSTALL_PREFIX=../dist \
-DCMAKE_PREFIX_PATH=/usr/llvm/18.1.8/dist \
-DCUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda-12.3.2 \
-DCUDA_DNN_ROOT_DIR=/usr/local/cudnn-9.2.1 \
-DCUDA_DNN_ROOT_DIR=/usr/local/cudnn-9.3.0 \
-DLLVM_EXTERNAL_LIT=/usr/llvm/18.1.8/build/bin/llvm-lit \
../hipify
Expand Down Expand Up @@ -696,7 +696,7 @@ The corresponding successful output is:
-- Initial CUDA to configure:
-- - CUDA Toolkit path : /usr/local/cuda-12.3.2
-- - CUDA Samples path :
-- - cuDNN path : /usr/local/cudnn-9.2.1
-- - cuDNN path : /usr/local/cudnn-9.3.0
-- - CUB path :
-- Found CUDAToolkit: /usr/local/cuda-12.3.2/targets/x86_64-linux/include (found version "12.3.107")
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
Expand All @@ -705,7 +705,7 @@ The corresponding successful output is:
-- Found CUDA config:
-- - CUDA Toolkit path : /usr/local/cuda-12.3.2
-- - CUDA Samples path : OFF
-- - cuDNN path : /usr/local/cudnn-9.2.1
-- - cuDNN path : /usr/local/cudnn-9.3.0
-- - CUB path : /usr/local/cuda-12.3.2/include/cub
-- Configuring done (0.5s)
-- Generating done (0.0s)
Expand Down Expand Up @@ -818,13 +818,13 @@ Tested configurations:
- ``3.11.4``
* - ``17.0.1`` :sup:`6` - ``18.1.8`` :sup:`7`
- ``7.0 - 12.3.2``
- ``8.0.5 - 9.2.1``
- ``8.0.5 - 9.3.0``
- ``2019.16.11.38, 2022.17.10.5``
- ``3.30.0``
- ``3.12.4``
* - ``19.0.0git``
- ``7.0 - 12.5.1``
- ``8.0.5 - 9.2.1``
- ``8.0.5 - 9.3.0``
- ``2019.16.11.38, 2022.17.10.5``
- ``3.30.0``
- ``3.12.4``
Expand Down Expand Up @@ -855,7 +855,7 @@ Building with testing support using ``Visual Studio 17 2022`` on ``Windows 11``:
-DCMAKE_PREFIX_PATH=D:/LLVM/18.1.8/dist \
-DCUDA_TOOLKIT_ROOT_DIR="C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.3" \
-DCUDA_SDK_ROOT_DIR="C:/ProgramData/NVIDIA Corporation/CUDA Samples/v12.3" \
-DCUDA_DNN_ROOT_DIR=D:/CUDA/cuDNN/9.2.1 \
-DCUDA_DNN_ROOT_DIR=D:/CUDA/cuDNN/9.3.0 \
-DLLVM_EXTERNAL_LIT=D:/LLVM/18.1.8/build/Release/bin/llvm-lit.py \
../hipify
Expand Down Expand Up @@ -892,13 +892,13 @@ The corresponding successful output is:
-- Initial CUDA to configure:
-- - CUDA Toolkit path : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.3
-- - CUDA Samples path : C:/ProgramData/NVIDIA Corporation/CUDA Samples/v12.3
-- - cuDNN path : D:/CUDA/cuDNN/9.2.1
-- - cuDNN path : D:/CUDA/cuDNN/9.3.0
-- - CUB path :
-- Found CUDAToolkit: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.3/include (found version "12.3.107")
-- Found CUDA config:
-- - CUDA Toolkit path : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.3
-- - CUDA Samples path : C:/ProgramData/NVIDIA Corporation/CUDA Samples/v12.3
-- - cuDNN path : D:/CUDA/cuDNN/9.2.1
-- - cuDNN path : D:/CUDA/cuDNN/9.3.0
-- - CUB path : C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.3/include/cub
-- Configuring done (1.4s)
-- Generating done (0.1s)
Expand Down
2 changes: 1 addition & 1 deletion docs/tables/CUDNN_API_supported_by_HIP.md
Original file line number Diff line number Diff line change
Expand Up @@ -886,7 +886,7 @@
|`cudnnBackendExecute`|8.0.1| | | | | | | | | |
|`cudnnBackendFinalize`|8.0.1| | | | | | | | | |
|`cudnnBackendGetAttribute`|8.0.1| | | | | | | | | |
|`cudnnBackendInitialize`|8.0.1| | | | | | | | | |
|`cudnnBackendInitialize`|8.0.1|9.3.0| | | | | | | | |
|`cudnnBackendSetAttribute`|8.0.1| | | | | | | | | |
|`cudnnBatchNormalizationBackward`|4.0.0|9.0.0| | |`hipdnnBatchNormalizationBackward`| | | | | |
|`cudnnBatchNormalizationBackwardEx`|7.4.1|9.0.0| | | | | | | | |
Expand Down
2 changes: 1 addition & 1 deletion docs/tables/CUDNN_API_supported_by_HIP_and_MIOPEN.md
Original file line number Diff line number Diff line change
Expand Up @@ -886,7 +886,7 @@
|`cudnnBackendExecute`|8.0.1| | | | | | | | | | | | | | | |
|`cudnnBackendFinalize`|8.0.1| | | | | | | | | | | | | | | |
|`cudnnBackendGetAttribute`|8.0.1| | | | | | | | | | | | | | | |
|`cudnnBackendInitialize`|8.0.1| | | | | | | | | | | | | | | |
|`cudnnBackendInitialize`|8.0.1|9.3.0| | | | | | | | | | | | | | |
|`cudnnBackendSetAttribute`|8.0.1| | | | | | | | | | | | | | | |
|`cudnnBatchNormalizationBackward`|4.0.0|9.0.0| | |`hipdnnBatchNormalizationBackward`| | | | | |`miopenBatchNormalizationBackward`| | | | | |
|`cudnnBatchNormalizationBackwardEx`|7.4.1|9.0.0| | | | | | | | | | | | | | |
Expand Down
2 changes: 1 addition & 1 deletion docs/tables/CUDNN_API_supported_by_MIOPEN.md
Original file line number Diff line number Diff line change
Expand Up @@ -886,7 +886,7 @@
|`cudnnBackendExecute`|8.0.1| | | | | | | | | |
|`cudnnBackendFinalize`|8.0.1| | | | | | | | | |
|`cudnnBackendGetAttribute`|8.0.1| | | | | | | | | |
|`cudnnBackendInitialize`|8.0.1| | | | | | | | | |
|`cudnnBackendInitialize`|8.0.1|9.3.0| | | | | | | | |
|`cudnnBackendSetAttribute`|8.0.1| | | | | | | | | |
|`cudnnBatchNormalizationBackward`|4.0.0|9.0.0| | |`miopenBatchNormalizationBackward`| | | | | |
|`cudnnBatchNormalizationBackwardEx`|7.4.1|9.0.0| | | | | | | | |
Expand Down
2 changes: 1 addition & 1 deletion src/CUDA2HIP_DNN_API_functions.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -432,7 +432,7 @@ const std::map<llvm::StringRef, cudaAPIversions> CUDA_DNN_FUNCTION_VER_MAP {
{"cudnnAdvTrainVersionCheck", {CUDNN_801, CUDA_0, CUDNN_900}},
{"cudnnBackendCreateDescriptor", {CUDNN_801, CUDA_0, CUDA_0 }},
{"cudnnBackendDestroyDescriptor", {CUDNN_801, CUDA_0, CUDA_0 }},
{"cudnnBackendInitialize", {CUDNN_801, CUDA_0, CUDA_0 }},
{"cudnnBackendInitialize", {CUDNN_801, CUDNN_930, CUDA_0 }},
{"cudnnBackendFinalize", {CUDNN_801, CUDA_0, CUDA_0 }},
{"cudnnBackendSetAttribute", {CUDNN_801, CUDA_0, CUDA_0 }},
{"cudnnBackendGetAttribute", {CUDNN_801, CUDA_0, CUDA_0 }},
Expand Down
1 change: 1 addition & 0 deletions src/Statistics.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -541,6 +541,7 @@ std::string Statistics::getCudaVersion(const cudaVersions &ver) {
case CUDNN_900: return "9.0.0";
case CUDNN_910: return "9.1.0";
case CUDNN_920: return "9.2.0";
case CUDNN_930: return "9.3.0";
}
return "";
}
Expand Down
3 changes: 2 additions & 1 deletion src/Statistics.h
Original file line number Diff line number Diff line change
Expand Up @@ -295,7 +295,8 @@ enum cudaVersions {
CUDNN_900 = 900,
CUDNN_910 = 910,
CUDNN_920 = 920,
CUDNN_LATEST = CUDNN_920,
CUDNN_930 = 930,
CUDNN_LATEST = CUDNN_930,
};

enum hipVersions {
Expand Down

0 comments on commit 67c2626

Please sign in to comment.