Robert Jacob

So with ne20 there's no performance degradation? Does the M100 start to outperform the V100 if you try a larger problem?

merged to next.

Why is it non-bfb even when chemistry is not active?

O3v2 was added in https://github.com/E3SM-Project/E3SM/pull/3047. You're saying you had to alter that in a non-bfb way to work with atmospheric chemistry? Does "linoz v2" refer to the linoz algorithm in...

The workstation results are bfb with cori-haswell?

Yes just merge it.

@glemieux no you don't have to compare to a known machine.

First you have to add his fork as a remote in your clone. That's covered in the Integrators guide under "Integrating from a fork" https://acme-climate.atlassian.net/wiki/spaces/DOC/pages/1708589129/Integrator+Guide#IntegratorGuide-Integratingfromafork

Yes since you now have 2 remotes in your clone.

Nothing is wrong. Do what it said to get "next".