# Is it possible to use GPU to train graphSage?

**URL:** <https://community.neo4j.com/t/is-it-possible-to-use-gpu-to-train-graphsage/38872>\
**Category:** Neo4j Graph Platform\
**Created:** [May 26, 2021, 1:00am UTC](https://community.neo4j.com/t/is-it-possible-to-use-gpu-to-train-graphsage/38872 "2021-05-26T01:00:15Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![lingvisa](https://sea1.discourse-cdn.com/flex021/user_avatar/community.neo4j.com/lingvisa/32/8150_2.png) [@lingvisa](https://community.neo4j.com/u/lingvisa)\
**Post date:** [May 26, 2021, 1:00am UTC](https://community.neo4j.com/t/is-it-possible-to-use-gpu-to-train-graphsage/38872/1 "2021-05-26T01:00:15Z")

</div>

My graph needs about 2 hours to train graphSage algorithm once and I have access to GPU. Is it possible to configure the graphSage algorithm to take advantage of GPUs? My typical use case is to train on a single machine, which could have multiple GPUs.

Thanks for advice.