# GPU mem. Occupied Too Much

**URL:** <https://forums.developer.nvidia.com/t/gpu-mem-occupied-too-much/75316>\
**Category:** Jetson Nano\
**Created:** [May 24, 2019, 11:23am UTC](https://forums.developer.nvidia.com/t/gpu-mem-occupied-too-much/75316 "2019-05-24T11:23:24Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![orcdnz](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@orcdnz](https://forums.developer.nvidia.com/u/orcdnz)\
**Post date:** [May 24, 2019, 11:23am UTC](https://forums.developer.nvidia.com/t/gpu-mem-occupied-too-much/75316/1 "2019-05-24T11:23:24Z")

</div>

Hey,

When I run object detection using the GPU on my Nano, it says;

totalMemory:3.86 GiB freeMemory:622.61MiB

Im not running anything else on the GPU, does anyone know why is this happening and how can I can increase the free memory for the operation?

---

<div class="post-metadata">

**Author:** ![AastaLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/aastalll/32/14043_2.png) [@AastaLLL](https://forums.developer.nvidia.com/u/AastaLLL)\
**Post date:** [May 27, 2019, 3:03am UTC](https://forums.developer.nvidia.com/t/gpu-mem-occupied-too-much/75316/2 "2019-05-27T03:03:05Z")

</div>

Hi,

Suppose this is a TensorFlow log.  
Could you try to restart the Nano and run the program again?  
TensorFlow may not correctly capture the memory size sometimes on the system that sharing memory between CPU and GPU.

Thanks.