I use NVIDIA GeForce 1050 Ti.
I have installed CUDA Toolkit 10.0 and [ cuDNN v7.6.5 (November 5th, 2019), for CUDA 10.0]
When I run tf.test.is_gpu_available() it returns False.
print(tf.test.is_built_with_cuda()) returns True.
What is it that I am missing?
erw
2
Hello,
I’m having the same issue with *tf.test.is_gpu_available() and tf.test.is_built_with_cuda(). Did you have any luck?
Cheers
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| torch.cuda.is_available() returns False | 0 | 1470 | April 19, 2019 | |
| RTX 3070, Cuda and Tensorflow version | 0 | 1383 | December 5, 2020 | |
| Tensorflow doesn't use GPU | 4 | 2572 | March 26, 2025 | |
| TensorFlow says CUDA not enabled | 3 | 8337 | June 21, 2019 | |
| CUDA 10.2 & Tensorflow 2.0. Getting an error when testing Tensorflow | 7 | 21126 | March 20, 2020 | |
| NVIDIA GeForce 940MX torch.cuda.is_available() False | 0 | 918 | July 20, 2020 | |
| 1660 Cuda compatibility | 1 | 1497 | December 12, 2020 | |
| cudaGetDeviceCount returned 100 -> no CUDA-capable device is detected | 1 | 1711 | June 6, 2023 | |
| Trying to implement GPU for Tensorflow using CUDA & cuDNN | 1 | 1213 | May 20, 2024 | |
| GPU not detected with Tensorflow | 1 | 810 | May 4, 2025 |