CODEWITHSUNDEEP
pythontensorflow
gazm2k5
gazm2k5

Reputation: 490

TensorFlow-GPU not finding GPU

I have installed tensorflow-gpu 1.15.2 on a Python 3.7 venv. I do not have the tensorflow package installed.

I installed CUDA 9.0 (as I'm using tensorflow 1.15) and the corresponding cuDNN for CUDA 9.0. When I run a tensorflow learning algorithm, it uses my CPU instead of my GPU. I ran:

from tensorflow.python.client import device_lib
print(device_lib.list_local_devices())

And it detects my GPU (1660 Ti) but it says I am missing loads of packages:

2020-04-25 22:02:12.536321: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2020-04-25 22:02:15.175536: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2020-04-25 22:02:15.188183: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library nvcuda.dll
2020-04-25 22:02:15.234070: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1639] Found device 0 with properties: 
name: GeForce GTX 1660 Ti major: 7 minor: 5 memoryClockRate(GHz): 1.875
pciBusID: 0000:27:00.0
2020-04-25 22:02:15.239530: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'cudart64_100.dll'; dlerror: cudart64_100.dll not found
2020-04-25 22:02:15.242919: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'cublas64_100.dll'; dlerror: cublas64_100.dll not found
2020-04-25 22:02:15.251483: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'cufft64_100.dll'; dlerror: cufft64_100.dll not found
2020-04-25 22:02:15.255358: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'curand64_100.dll'; dlerror: curand64_100.dll not found
2020-04-25 22:02:15.266446: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'cusolver64_100.dll'; dlerror: cusolver64_100.dll not found
ired libraries for your platform.
Skipping registering GPU devices...
2020-04-25 22:02:15.719511: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1180] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-04-25 22:02:15.721901: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1186]      0
2020-04-25 22:02:15.723610: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1199] 0:   N
[name: "/device:CPU:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 524594082372294943
]

I did a search on my PC and indeed can't find those dll files.

When I installed CUDA, and if I try to reinstall, it says "The graphics driver could not find compatible graphics hardware." I ignored the error and installed anyway. The relevant CUDA directories in my Program Files are there. But nvcc -V in cmd prompt confirms that CUDA is installed.

A quick google seemed to suggest CUDA version 10+ has these libraries, but after installing I've seen no difference.

Where do I find these missing dlls? Have I installed something incorrectly?

Upvotes: 1

Views: 1083

Answers (1)

gazm2k5
gazm2k5

Reputation: 490

Oh it seems the 100 refers to CUDA 10.0, so CUDA 9.0 has ***90.dll files and CUDA 10.2 has ***102.dll files.

Really annoying to need so many different installations of CUDA for this!

Upvotes: 3

Related Questions