# Can I use AN AMD or Nvidia graphics card to speed up MNIST training and recognition?

**URL:** <https://community.konduit.ai/t/can-i-use-an-amd-or-nvidia-graphics-card-to-speed-up-mnist-training-and-recognition/1053>\
**Category:** Uncategorized\
**Created:** [December 11, 2020, 4:13pm UTC](https://community.konduit.ai/t/can-i-use-an-amd-or-nvidia-graphics-card-to-speed-up-mnist-training-and-recognition/1053 "2020-12-11T16:13:48Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![TRSK-CONMI](https://yyz1.discourse-cdn.com/flex035/user_avatar/community.konduit.ai/trsk-conmi/32/368_2.png) [@TRSK-CONMI](https://community.konduit.ai/u/TRSK-CONMI)\
**Post date:** [December 11, 2020, 4:13pm UTC](https://community.konduit.ai/t/can-i-use-an-amd-or-nvidia-graphics-card-to-speed-up-mnist-training-and-recognition/1053/1 "2020-12-11T16:13:48Z")

</div>

Any examples of that?

---

<div class="post-metadata">

**Author:** ![treo](https://yyz1.discourse-cdn.com/flex035/user_avatar/community.konduit.ai/treo/32/47_2.png) [@treo](https://community.konduit.ai/u/treo)\
**Post date:** [December 11, 2020, 5:04pm UTC](https://community.konduit.ai/t/can-i-use-an-amd-or-nvidia-graphics-card-to-speed-up-mnist-training-and-recognition/1053/2 "2020-12-11T17:04:45Z")

</div>

Models that work on MNIST are usually very small. Even the entire mnist dataset is less than 50MB even when uncompressed.

GPU’s usually excel when they are used with large models and data sizes, as otherwise they usually are bottlenecked by the PCIe communication latency.

Because MNIST data is so small and the models used on it are equally small, you will typically need to use very large batch sizes to see any benefit at all when using a GPU.

But in principle you can take any MNIST example (see [deeplearning4j-examples/dl4j-examples at master · eclipse/deeplearning4j-examples · GitHub](https://github.com/eclipse/deeplearning4j-examples/tree/master/dl4j-examples)) and change to the CUDA backend in the pom.xml (see [deeplearning4j-examples/pom.xml at master · eclipse/deeplearning4j-examples · GitHub](https://github.com/eclipse/deeplearning4j-examples/blob/master/dl4j-examples/pom.xml#L32-L33)) , and you will be ready to use a NVidia GPU.

As for AMD GPUs, those are currently not supported.

---

<div class="post-metadata">

**Author:** ![TRSK-CONMI](https://yyz1.discourse-cdn.com/flex035/user_avatar/community.konduit.ai/trsk-conmi/32/368_2.png) [@TRSK-CONMI](https://community.konduit.ai/u/TRSK-CONMI)\
**Post date:** [December 12, 2020, 3:45am UTC](https://community.konduit.ai/t/can-i-use-an-amd-or-nvidia-graphics-card-to-speed-up-mnist-training-and-recognition/1053/3 "2020-12-12T03:45:48Z")

</div>

Thank you,treo 🤝🤝🤝