# Utilise the GPU?

**URL:** <https://forum.cursor.com/t/utilise-the-gpu/155926>\
**Category:** Feature Requests\
**Tags:** performance\
**Created:** [March 26, 2026, 8:57am UTC](https://forum.cursor.com/t/utilise-the-gpu/155926 "2026-03-26T08:57:05Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![ryan\_clarke](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/ryan_clarke/32/103226_2.png) [@ryan\_clarke](https://forum.cursor.com/u/ryan_clarke)\
**Post date:** [March 26, 2026, 8:57am UTC](https://forum.cursor.com/t/utilise-the-gpu/155926/1 "2026-03-26T08:57:05Z")

</div>

### Feature request for product/service

Cursor IDE

### Describe the request

It might not be possible but I was playing with Illama using a custom model and it is so fast.  
I also know the GPU has memory limits  
I was wondering if Cursor could have a hybrid approach and use the GPU for chat/organising things and doing simple tasks which the remote LLM does the deep thinking.  
This could speed things up significantly if it worked and allowed us to use deep thinking a lot more without breaking the bank

### Operating System (if it applies)

Windows 10/11