# For MIG on A100 and A30, can different GPU instance communicate?

**URL:** <https://forums.developer.nvidia.com/t/for-mig-on-a100-and-a30-can-different-gpu-instance-communicate/238387>\
**Category:** CUDA Programming and Performance\
**Created:** [January 2, 2023, 7:16am UTC](https://forums.developer.nvidia.com/t/for-mig-on-a100-and-a30-can-different-gpu-instance-communicate/238387 "2023-01-02T07:16:47Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![202476410arsmart](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@202476410arsmart](https://forums.developer.nvidia.com/u/202476410arsmart)\
**Post date:** [January 2, 2023, 7:16am UTC](https://forums.developer.nvidia.com/t/for-mig-on-a100-and-a30-can-different-gpu-instance-communicate/238387/1 "2023-01-02T07:16:47Z")

</div>

Hi! I learnt that in A100 exist a “MIG” technique to divide one GPU into separate small GPUs, and I am wondering can they communicate with each other? Thanks!!

> **[NVIDIA Multi-Instance GPU(MIG)](https://www.nvidia.com/en-us/technologies/multi-instance-gpu/)**
>
> Expand GPU Access to More Users and Flexibility for Every Workload.

---

<div class="post-metadata">

**Author:** ![Robert\_Crovella](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/robert_crovella/32/14043_2.png) [@Robert\_Crovella](https://forums.developer.nvidia.com/u/Robert_Crovella)\
**Post date:** [January 2, 2023, 4:41pm UTC](https://forums.developer.nvidia.com/t/for-mig-on-a100-and-a30-can-different-gpu-instance-communicate/238387/2 "2023-01-02T16:41:32Z")

</div>

Ordinary communications like host-based MPI are possible. Also, a limited (between CI but not between GI) form of CUDA IPC [is possible](https://docs.nvidia.com/datacenter/tesla/mig-user-guide/index.html#cuda-concurrency). P2P is [not possible between MIG partitions](https://docs.nvidia.com/datacenter/tesla/mig-user-guide/index.html#app-considerations), currently.