# Cudnn\_status\_not\_initialized

**URL:** <https://discuss.pytorch.org/t/cudnn-status-not-initialized/7090>\
**Category:** Uncategorized\
**Created:** [September 6, 2017, 3:15am UTC](https://discuss.pytorch.org/t/cudnn-status-not-initialized/7090 "2017-09-06T03:15:09Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![zhengyun](https://discuss.pytorch.org/letter_avatar_proxy/v4/letter/z/48db29/32.png) [@zhengyun](https://discuss.pytorch.org/u/zhengyun)\
**Post date:** [September 6, 2017, 3:15am UTC](https://discuss.pytorch.org/t/cudnn-status-not-initialized/7090/1 "2017-09-06T03:15:09Z")

</div>

The error is shown below:

File “/home/zhengyun.zy/anaconda/lib/python2.7/site-packages/torch/nn/modules/module.py”, line 206, in **call**  
result = self.forward(\*input, \*\*kwargs)  
File “/home/zhengyun.zy/anaconda/lib/python2.7/site-packages/torch/nn/modules/conv.py”, line 237, in forward  
self.padding, self.dilation, self.groups)  
File “/home/zhengyun.zy/anaconda/lib/python2.7/site-packages/torch/nn/functional.py”, line 40, in conv2d  
return f(input, weight, bias)  
RuntimeError: CUDNN\_STATUS\_NOT\_INITIALIZED

cudnn is correctly installed

> > > import torch  
> > > print(torch.backends.cudnn.is\_acceptable(torch.cuda.FloatTensor(1)))  
> > > True  
> > > print(torch.backends.cudnn.version())  
> > > 6021

---

<div class="post-metadata">

**Author:** ![ngimel](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/ngimel/32/14766_2.png) [@ngimel](https://discuss.pytorch.org/u/ngimel)\
**Post date:** [September 6, 2017, 6:43pm UTC](https://discuss.pytorch.org/t/cudnn-status-not-initialized/7090/2 "2017-09-06T18:43:41Z")

</div>

According to the manual, this error is returned when no compatible GPU is found, CUDA driver is not installed or disabled, CUDA runtime API initialization failed.

---

<div class="post-metadata">

**Author:** ![Rizhiy](https://discuss.pytorch.org/user_avatar/discuss.pytorch.org/rizhiy/32/18854_2.png) [@Rizhiy](https://discuss.pytorch.org/u/Rizhiy)\
**Post date:** [January 21, 2018, 5:54pm UTC](https://discuss.pytorch.org/t/cudnn-status-not-initialized/7090/3 "2018-01-21T17:54:01Z")

</div>

Any update on this? I’m experiencing the same problem with CUDA 9.1 and cuDNN 7

---

<div class="post-metadata">

**Author:** ![Maksim\_Kretov](https://discuss.pytorch.org/letter_avatar_proxy/v4/letter/m/ad7895/32.png) [@Maksim\_Kretov](https://discuss.pytorch.org/u/Maksim_Kretov)\
**Post date:** [January 27, 2018, 2:16pm UTC](https://discuss.pytorch.org/t/cudnn-status-not-initialized/7090/4 "2018-01-27T14:16:58Z")

</div>

Had the same issue, resolved after I replaced cuda 9.1 with cuda 9.0.  
Cudnn 7, pytorch 0.3.