# CPU/GPU node, job schedule

**URL:** <https://community.openpbs.org/t/cpu-gpu-node-job-schedule/3020>\
**Category:** Users/Site Administrators\
**Created:** [March 1, 2022, 1:00pm UTC](https://community.openpbs.org/t/cpu-gpu-node-job-schedule/3020 "2022-03-01T13:00:15Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Juergen](https://avatars.discourse-cdn.com/v4/letter/j/65b543/32.png) [@Juergen](https://community.openpbs.org/u/Juergen)\
**Post date:** [March 1, 2022, 1:00pm UTC](https://community.openpbs.org/t/cpu-gpu-node-job-schedule/3020/1 "2022-03-01T13:00:15Z")

</div>

Hello, I have the following problem:

Node1: 128 CPU cores  
Node2: 64 CPU Cores/4GPU

Now all CPU’s of Node2 are busy because jobs were sent that also need GPUs while almost all CPU’s on Node1 are free. Is it possible to define a QUEUE in such a way that the job uses the GPU’s of Node 2 but the CPU’s of Node1?

thanks and best wishes;  
Juergen

---

<div class="post-metadata">

**Author:** ![adarsh](https://avatars.discourse-cdn.com/v4/letter/a/f07891/32.png) [@adarsh](https://community.openpbs.org/u/adarsh)\
**Post date:** [March 1, 2022, 4:06pm UTC](https://community.openpbs.org/t/cpu-gpu-node-job-schedule/3020/2 "2022-03-01T16:06:40Z")

</div>

> [@Juergen](#):
>
> Is it possible to define a QUEUE in such a way that the job uses the GPU’s of Node 2 but the CPU’s of Node1?

No, it is not possible.

- QUEUE is a place holder for jobs , it does not make any decisions.
- without using openPBS, does your applicaiton support such a usage of 2 cpus of node1 and gpus of node2 ?