Facing error while importing RAPIDS library in H100 GPU environment

AI & Data Science Deep Learning (Training & Inference) TensorRT
, , , ,
1

I am trying to run RAPIDS library in my H100 GPU.

I tried installing RAPIDS in my environment using this: RAPIDS | GPU Accelerated Data Science*

But while importing cudf library, am getting the following error:

In [1]: import cudf
/home/nvidia/.local/lib/python3.10/site-packages/cudf/utils/_ptxcompiler.py:61: UserWarning: Error getting driver and runtime versions:

stdout:



stderr:

Traceback (most recent call last):
  File "/home/nvidia/.local/lib/python3.10/site-packages/numba/cuda/cudadrv/driver.py", line 258, in ensure_initialized
    self.cuInit(0)
  File "/home/nvidia/.local/lib/python3.10/site-packages/numba/cuda/cudadrv/driver.py", line 331, in safe_cuda_api_call
    self._check_ctypes_error(fname, retcode)
  File "/home/nvidia/.local/lib/python3.10/site-packages/numba/cuda/cudadrv/driver.py", line 399, in _check_ctypes_error
    raise CudaAPIError(retcode, msg)
numba.cuda.cudadrv.driver.CudaAPIError: [101] Call to cuInit results in CUDA_ERROR_INVALID_DEVICE

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<string>", line 4, in <module>
  File "/home/nvidia/.local/lib/python3.10/site-packages/numba/cuda/cudadrv/driver.py", line 296, in __getattr__
    self.ensure_initialized()
  File "/home/nvidia/.local/lib/python3.10/site-packages/numba/cuda/cudadrv/driver.py", line 262, in ensure_initialized
    raise CudaSupportError(f"Error at driver init: {description}")
numba.cuda.cudadrv.error.CudaSupportError: Error at driver init: Call to cuInit results in CUDA_ERROR_INVALID_DEVICE (101)


Not patching Numba
  warnings.warn(msg, UserWarning)
---------------------------------------------------------------------------
CUDARuntimeError                          Traceback (most recent call last)
<ipython-input-1-e13365c50bc4> in <module>
----> 1 import cudf

~/.local/lib/python3.10/site-packages/cudf/__init__.py in <module>
      8 
      9 _setup_numba()
---> 10 validate_setup()
     11 
     12 import cupy

~/.local/lib/python3.10/site-packages/cudf/utils/gpu_utils.py in validate_setup()
     53     except CUDARuntimeError as e:
     54         if e.status in notify_caller_errors:
---> 55             raise e
     56         # If there is no GPU detected, set `gpus_count` to -1
     57         gpus_count = -1

~/.local/lib/python3.10/site-packages/cudf/utils/gpu_utils.py in validate_setup()
     50 
     51     try:
---> 52         gpus_count = getDeviceCount()
     53     except CUDARuntimeError as e:
     54         if e.status in notify_caller_errors:

~/.local/lib/python3.10/site-packages/rmm/_cuda/gpu.py in getDeviceCount()
    100     status, count = cudart.cudaGetDeviceCount()
    101     if status != cudart.cudaError_t.cudaSuccess:
--> 102         raise CUDARuntimeError(status)
    103     return count
    104 

CUDARuntimeError: cudaErrorInvalidDevice: invalid device ordinal
3

Hi @techgig ,
I believe you can raise the issue on Issues - RAPIDS Docs

Thanks