# Recommendations for -np0 and -np1

**URL:** <https://discourse.channelflow.ch/t/recommendations-for-np0-and-np1/33>\
**Category:** Usage\
**Created:** [October 10, 2018, 1:58pm UTC](https://discourse.channelflow.ch/t/recommendations-for-np0-and-np1/33 "2018-10-10T13:58:36Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![john.gibson](https://yyz1.discourse-cdn.com/flex027/user_avatar/discourse.channelflow.ch/john.gibson/32/11_2.png) [@john.gibson](https://discourse.channelflow.ch/u/john.gibson)\
**Post date:** [October 10, 2018, 1:58pm UTC](https://discourse.channelflow.ch/t/recommendations-for-np0-and-np1/33/1 "2018-10-10T13:58:36Z")

</div>

I’m running channelflow on a cluster with 28 compute nodes. Each node has two 12-core Xeon E5-2680 CPUS.

I envision two scales of channelflow computations:

- small, running on one node with 24 cores, 128 x 129 x 128 discretization
- large, running on, say, sixteen nodes with 384 cores, 1024 x 129 x 1024 discretization

What are decent starting point for values of `-np0` and `-np1` for these two simulations?

---

<div class="post-metadata">

**Author:** ![florian.reetz](https://yyz1.discourse-cdn.com/flex027/user_avatar/discourse.channelflow.ch/florian.reetz/32/12_2.png) [@florian.reetz](https://discourse.channelflow.ch/u/florian.reetz)\
**Post date:** [October 10, 2018, 3:10pm UTC](https://discourse.channelflow.ch/t/recommendations-for-np0-and-np1/33/2 "2018-10-10T15:10:52Z")

</div>

Hello John,

there is only one hard requirement for the MPI distribution which is `mod(Nx,np0)=0`. The reason is the way the FFTW transform plans are set up. If you do not respect this, you will see a runtime error. In your case, any power of 2 will work for `np0`.

I typically use a distribution which is close to equally distributed because makes the data chunks most likely to be the same on each process (“pencil distribution”). In your case, this would be `(np0,np1)=(4,6)` for small, and `(np0,np1)=(16,24)` for large. However, it might be worth to test if a “slab distribution” is faster in your case. This means that only one dimension is distributed, let’s say `(np0,np1)=(1,384)` for large. The advantage is that you save cost for communication. The disadvantage is that it is more likely to distribute the data unequally over the tasks. np1 divides the x-dimension in physical space and the z-dimension in spectral. If your numerical domain is very long but narrow, a “slab distribution” is probably a bad idea but your domain has equal sides and it might be more performant.

---

<div class="post-metadata">

**Author:** ![laurette](https://avatars.discourse-cdn.com/v4/letter/l/e99b99/32.png) [@laurette](https://discourse.channelflow.ch/u/laurette)\
**Post date:** [May 27, 2019, 9:23am UTC](https://discourse.channelflow.ch/t/recommendations-for-np0-and-np1/33/3 "2019-05-27T09:23:57Z")

</div>

On the computer and system that we use (IBM x3750) we find that (np0,np1)=(1,np) is optimal for our (Lx,Lz)=(10,40) domains, at least up to np=64.

---

<div class="post-metadata">

**Author:** ![john.gibson](https://yyz1.discourse-cdn.com/flex027/user_avatar/discourse.channelflow.ch/john.gibson/32/11_2.png) [@john.gibson](https://discourse.channelflow.ch/u/john.gibson)\
**Post date:** [September 5, 2019, 4:00pm UTC](https://discourse.channelflow.ch/t/recommendations-for-np0-and-np1/33/4 "2019-09-05T16:00:39Z")

</div>

Laurette: I experience the same on my Intel Xeon CPU E5-2680 cluster. (np0, np1) = (1,np) is optimal for all np I’ve tested, up to 64.