# cudnnGetConvolutionForwardWorkspaceSize - possible overflow?

**URL:** <https://forums.developer.nvidia.com/t/cudnngetconvolutionforwardworkspacesize-possible-overflow/36550>\
**Category:** GPU-Accelerated Libraries\
**Created:** [January 28, 2015, 9:10pm UTC](https://forums.developer.nvidia.com/t/cudnngetconvolutionforwardworkspacesize-possible-overflow/36550 "2015-01-28T21:10:27Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Koobas](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@Koobas](https://forums.developer.nvidia.com/u/Koobas)\
**Post date:** [January 28, 2015, 9:10pm UTC](https://forums.developer.nvidia.com/t/cudnngetconvolutionforwardworkspacesize-possible-overflow/36550/1 "2015-01-28T21:10:27Z")

</div>

I am trying to run an example from the paper “cuDNN: Efficient Primitives for Deep Learning”.  
Everything seems to be in order, but the function cudnnGetConvolutionForwardWorkspaceSize seems to have a problem.  
For instance, if I set the parameters to those of Layer 2 of Table 2, e.g. (N, C, H, W, K, R, S) = (128, 96, 64, 64, 128, 9, 9), then the function returns an insane number.  
If I just change the first number, the batch size, from 128 to 64, the function returns a reasonable 5.8GB, but at 128 it returns a number that is obviously wrong.  
I am not excluding the possibility that the error is in my code, however, I think there is a chance that overflow happens in that function.  
Jakub

---

<div class="post-metadata">

**Author:** ![adshen](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@adshen](https://forums.developer.nvidia.com/u/adshen)\
**Post date:** [February 5, 2015, 4:05am UTC](https://forums.developer.nvidia.com/t/cudnngetconvolutionforwardworkspacesize-possible-overflow/36550/2 "2015-02-05T04:05:43Z")

</div>

It is a bug in cudnn. The fix will be available for you in the future release. Thank you for reporting.