build the master branch with cuda10.2/cudnn8.0

[cqray1990]

Describe the bug
A clear and concise description of what the bug is. To avoid repetition please make sure this is not one of the known issues mentioned on the respective release page.

Urgency
If there are particular important use cases blocked by this or strict project-related timelines, please share more information and dates. If there are no hard deadlines, please specify none.

System information

OS Platform and Distribution (e.g., Linux Ubuntu 16.04):
ONNX Runtime installed from (source or binary):
ONNX Runtime version:master branch
Python version:3.6
Visual Studio version (if applicable):
GCC/Compiler version (if compiling from source):
CUDA/cuDNN version:10.2/8.0
GPU model and memory:

To Reproduce

Describe steps/code to reproduce the behavior.
Attach the ONNX model to the issue (where applicable) to expedite investigation.

when execte ./build.sh --parallel --build_shared_lib --build_wheel --config=RelWithDebInfo --use_cuda --cuda_home /usr/local/cuda --cudnn_home /usr/local/cuda --cuda_version=10.2 --cmake_extra_defines CMAKE_CUDA_ARCHITECTURES=52 --skip_tests

adding 'onnxruntime_gpu-1.10.0.data/purelib/onnxruntime/transformers/longformer/longformer_helper.py'
adding 'onnxruntime_gpu-1.10.0.dist-info/METADATA'
adding 'onnxruntime_gpu-1.10.0.dist-info/WHEEL'
adding 'onnxruntime_gpu-1.10.0.dist-info/entry_points.txt'
adding 'onnxruntime_gpu-1.10.0.dist-info/top_level.txt'
adding 'onnxruntime_gpu-1.10.0.dist-info/RECORD'
removing build/bdist.linux-x86_64/wheel
2021-10-31 23:19:13,451 util.run [DEBUG] - Subprocess completed. Return code: 0
2021-10-31 23:19:13,451 build [INFO] - Build complete

but when execute ./build.sh --parallel --build_shared_lib --build_wheel --config=RelWithDebInfo --use_cuda --cuda_home /usr/local/cuda --cudnn_home /usr/local/cuda --cuda_version=10.2 --cmake_extra_defines CMAKE_CUDA_ARCHITECTURES=52 with no --skip_tests flag

errors comes:

6/6 Test #6: onnxruntime_api_tests_without_env ...... Passed 0.02 sec

83% tests passed, 1 tests failed out of 6

Total Test time (real) = 36.69 sec

The following tests FAILED:
1 - onnxruntime_test_all (Failed)
Errors while running CTest
Traceback (most recent call last):
File "sensors/onnxruntime/tools/ci_build/build.py", line 2353, in
sys.exit(main())
File "/sensors/onnxruntime/tools/ci_build/build.py", line 2277, in main
run_onnxruntime_tests(args, source_dir, ctest_path, build_dir, configs)
File "sensors/onnxruntime/tools/ci_build/build.py", line 1592, in run_onnxruntime_tests
run_subprocess(ctest_cmd, cwd=cwd, dll_path=dll_path)
File "sensors/onnxruntime/tools/ci_build/build.py", line 631, in run_subprocess
return run(*args, cwd=cwd, capture_stdout=capture_stdout, shell=shell, env=my_env)

[snnn]

Please redirect all the outputs to a text file and search "error:" within it.

[cqray1990]

@snnn cuda10.2 /cudnn8.0 2080ti drivers 460

[snnn]

See: https://developer.nvidia.com/cuda-gpus

You set wrong CMAKE_CUDA_ARCHITECTURES.

[cqray1990]

@snnn how set CMAKE_CUDA_ARCHITECTURES with 2080ti ,i install cuda 10.2 with cudnn 8.0 as official suggestion,what's the problem is?

[snnn]

Please find your card in https://developer.nvidia.com/cuda-gpus . Then set CMAKE_CUDA_ARCHITECTURES to the number (without dot).

[stale]

This issue has been automatically marked as stale due to inactivity and will be closed in 7 days if no further activity occurs. If further support is needed, please provide an update and/or more details.