Skip to content
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

Uses the array data structure for ParticleField and pullbacks for AD #13

Open
wants to merge 229 commits into
base: master
Choose a base branch
from
Open

Uses the array data structure for ParticleField and pullbacks for AD #13

wants to merge 229 commits into from

Conversation

cibinjoseph
Copy link
Member

This PR is to review the changes made and make sure everything looks good. This may not need to be merged into master immediately.

@rymanderson
add FLOWFMM compatibility functions
2dfb484
@rymanderson
prepare for workstation test
0f3923b
@rymanderson
remove DifferentialEquations dependency
19e4867
@rymanderson
add benchmark script
e2240f0
@rymanderson
use MVectors
3b41d05
@rymanderson
add more benchmarks to script
6271257
@rymanderson
update FMM
5ff91fc
@rymanderson
passing unit tests with FLOWFMM
2b99805
@rymanderson
replace abstract ::Function type definition with concrete parameteric…
bf0682c 
… type
@rymanderson
bug fix
b96b2e8
@rymanderson
made UJ type-stable (matching ExaFMM benchmarks)
e7edc2c
@rymanderson
bug fix
2e4de05
@rymanderson
prepare benchmark
97daa09
@rymanderson
prep new benchmark
189b009
@rymanderson
passing unit tests
40f856f
@rymanderson
make expansion_order consistent and prep benchmark
b31c96f
@rymanderson
prep benchmark with fmm
1caa44d
@rymanderson
specify julia-1.6
78bf00b
@rymanderson
allow changing pfield.Uinf by using less restrictive type
56d5d34
@rymanderson
bug fix
c106932
@rymanderson
remove duplicate eltype definition
69bcba8
@rymanderson
add nonzero_sigma parameter
2c8f93d
@rymanderson
change nonzero_sigma default =false
8b6271e
@rymanderson
initialize particle field with custom float type
d9b28e7
@rymanderson
merge conflict on workstation
1101eb6
@rymanderson
ForwardDiff compatibility
c163a52
@rymanderson
add custom_UJ function
ceeaac6
@rymanderson
working with panels
ad44787
@rymanderson
fix viscous and SFS syntax error
df9de1a
@rymanderson
update fields explicitly in remove_particle
4fdbf47
rymanderson and others added 30 commits July 12, 2024 13:18
@rymanderson
Add padding to target size to nearest multiple of 10 for efficient p,…
35dcf3e 
…q launch config
@rymanderson
Benchmarking
bfa671b
@rymanderson
Add max threads per block for Pascal
3caadbd
@rymanderson
update fmm.direct! syntax to use multiple target indices
be1036a
@rymanderson
backwards-compatibility for DynamicSFS convenience syntax
e5aa5e8
@rymanderson
Add sort function to divisors
b5db201
@rymanderson
merge conflict
9be1cb9
Fix tests for numerical useGPU
bf7de7c
Ignore runfiles and scripts
d494f76
Add max thread limits and padding in multiples of 32
35856cb
@cibinjoseph
Revert to target_index for cpu direct()
b910152
@rymanderson
Make direct_gpu!() use source_indices and target_indices and revert d…
ab6f5dc 
…irect!() to previous form
@rymanderson
Make compatible with new direct_gpu!() in FastMultipole
c7eff92
@rymanderson
Add GPU kernel for multiple gpus
398d35a
@rymanderson
Correct leapfrog example to use multiple gpus
5905f85
Bug fixes for multiple gpus
75b2f65
Use gpu array list to avoid variable scoping issue in multiple-gpu ke…
d8fbc94 
…rnel
Set useGPU to 2 as default
450baec
Reduce parsing of for loops. Break fast
e0acfd4
Fix for massive allocations in check_launch()
164d7be
Remove T=T in arguments inside kernel
bad13ac
Move toggle_sfs condition outside loops
971a15a
@rymanderson
Adapt warmup function so that it compiles on all available gpus
dbbf6ab
@BTV25
addition to fluiddomain to allow Uinf included
5456de9
@rymanderson
Perform self-interaction on one gpu for entire particle field when le…
a34bafd 
…af count is 8
@rymanderson
Add additional conditions for single self-interaction on gpu
0d5243d
Remove unnecessary view for GPU DtoH copy
a11c782
@rymanderson
Reduce register pressure by using Int32 indices
ca301de
Correct dimension of target_system particles in gpu kernel
9d211ea
@rymanderson
merge conflicts
cf387d4
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@rymanderson rymanderson rymanderson left review comments

@Eric-P-Green Eric-P-Green Awaiting requested review from Eric-P-Green

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

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
4 participants
@cibinjoseph @rymanderson @Eric-P-Green @BTV25