Skip to content

Openacc phase 2 #734

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
basava70 wants to merge 7 commits into
base: main
Choose a base branch
from
Open

Openacc phase 2 #734

basava70 wants to merge 7 commits into from

Conversation

@basava70
Copy link
Collaborator

@basava70 basava70 commented Jul 1, 2025 •
edited
Loading

  • Make diff_tracers_ale subroutine run completely on the GPU
    • diff_tracers_ale
      • diff_part_hor_redi
      • diff_ver_part_expl_ale
      • diff_ver_part_redi_expl
      • diff_ver_part_impl_ale
      • diff_part_bh
        (These are not touched as they are part of recom)
      • diff_ver_recom_expl
      • get_seawater_viscocity
      • get_particle_density
      • ballast
      • ver_sinking_recom
      • ver_sinking_recom_benthos
  • Run benchmarks and compare with Phase 1 PR Phase 1 OpenACC #732

suvarchal
suvarchal reviewed Jul 1, 2025
View reviewed changes
src/oce_ale_tracer.F90
!here update only those initialized in the init_tracers. (values, valuesAB, edge_up_dn_grad, ...)
!$ACC UPDATE DEVICE(tracers%data(tr_num)%values, tracers%data(tr_num)%valuesAB) &
!$ACC DEVICE(tracers%work%edge_up_dn_grad) !!&
!!$ACC UPDATE DEVICE(tracers%data(tr_num)%values, tracers%data(tr_num)%valuesAB, tracers%data(tr_num)%valuesold)
Copy link
Collaborator

@suvarchal suvarchal Jul 1, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

is this 227 meant to be comment?

Copy link
Collaborator Author

@basava70 basava70 Jul 1, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I opened this while phase1 benchmarks are running. I branched it out from Phase1. That's why it's commented out. But after my first commit, we can start the conversation. Please ignore this PR for now.

suvarchal
suvarchal reviewed Jul 1, 2025
View reviewed changes
@basava70 basava70 marked this pull request as draft July 1, 2025 13:44
@basava70 basava70 mentioned this pull request Jul 1, 2025
Merged
5 tasks
@basava70
Copy link
Collaborator Author

basava70 commented Jul 1, 2025

Can I select "Resolve" for your reviews @suvarchal ?

@basava70 basava70 requested a review from suvarchal July 1, 2025 14:37
@basava70 basava70 self-assigned this Jul 24, 2025
@basava70 basava70 marked this pull request as ready for review July 24, 2025 12:48
@basava70
Copy link
Collaborator Author

basava70 commented Jul 24, 2025 •
edited
Loading

Benchmarks for Phase 1 & Phase 2

Run Length Run Length Unit Steps per Day Phase1 (s) Phase2 (s)
1 d 32 104.20 234.06
1 m 32 3400.23 7226.26
1 y 32 >26400.0 >26400.0
2 y 32 NA NA

Due to the usage of many if loops in the cpu code, there is a massive reduction in the warp occupancy rate, which shot up the computation time.
I leave now for your reviews. Please let me know if you would like any changes.
Best

@JanStreffing JanStreffing added this to the FESOM 2.7 milestone Aug 19, 2025
@JanStreffing JanStreffing modified the milestones: FESOM 2.7, FESOM 2.8 Sep 8, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@JanStreffing JanStreffing Awaiting requested review from JanStreffing

@mandresm mandresm Awaiting requested review from mandresm

@suvarchal suvarchal Awaiting requested review from suvarchal

At least 1 approving review is required to merge this pull request.

Assignees

@basava70 basava70

Labels

None yet

Projects

None yet

Milestone

FESOM 2.8

Development

Successfully merging this pull request may close these issues.

4 participants

@basava70 @suvarchal @JanStreffing