search

src-d / kmcuda

Large scale K-means and K-nn implementation on NVIDIA GPU / CUDA
Other
783 stars 144 forks source link

Installing in V100 + Ubuntu #108

Open srakrn opened 4 years ago

srakrn commented 4 years ago

Hello, I am relatively new to this, so I am seeking for an answer to a very newbie question.

I am using a V100 GPU, and would like to install kmcuda for Python. All I did is a pip install with the specified CUDA path and architecture version:

CUDA_ARCH=70 CUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda pip install git+https://github.com/src-d/kmcuda.git#subdirectory=src

However, an attempt to run the command results into the following error:

>>> import numpy as np
>>> from sklearn.datasets import make_blobs
>>> from libKMCUDA import kmeans_cuda
>>> X, y = make_blobs(20000, 800, 10)
>>> X = X.astype(np.float32)
>>> km = kmeans_cuda(X, 10)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ValueError: No such CUDA device exists

Here is the output from nvidia-smi:

❯ nvidia-smi
Wed May 27 06:46:34 2020       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 440.33.01    Driver Version: 440.33.01    CUDA Version: 10.2     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  Tesla V100-SXM2...  On   | 00000000:00:1B.0 Off |                    0 |
| N/A   41C    P0    36W / 300W |     11MiB / 16160MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   1  Tesla V100-SXM2...  On   | 00000000:00:1C.0 Off |                    0 |
| N/A   40C    P0    37W / 300W |     11MiB / 16160MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   2  Tesla V100-SXM2...  On   | 00000000:00:1D.0 Off |                    0 |
| N/A   38C    P0    41W / 300W |     11MiB / 16160MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   3  Tesla V100-SXM2...  On   | 00000000:00:1E.0 Off |                    0 |
| N/A   43C    P0    42W / 300W |     11MiB / 16160MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+

I am uncertain whether this is the same as issue #75 or not, but believed it is not, since I can't even manage to get the library to see the CUDA device.

Thank you for your generous help in advance.