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

[WIP] Improve multi-GPU BFS performance #4619

Closed
wants to merge 137 commits into from
Closed

[WIP] Improve multi-GPU BFS performance #4619

wants to merge 137 commits into from

Conversation

seunghwak
Copy link
Contributor

No description provided.

seunghwak added 30 commits July 15, 2024 14:21
@seunghwak
add a create_graph_from_edgelist function that takes edge list in mul…
a0d1f01 
…tiple chunks
@seunghwak
update R-mat graph generators to generate edge list in multiple chunks
55513ae
@seunghwak
Merge branch 'branch-24.08' of https://github.com/rapidsai/cugraph in…
2163fd8 
…to enh_graph_creation
@seunghwak
fix build error
a9dfb92
@seunghwak
delete unused functions
e7b33ca
@seunghwak
fix build errors
27ea550
@seunghwak
add temporary performance measurement code
e5e8257
@seunghwak
add code to broadcast frontier using a bitmap
7ec5b08
@seunghwak
resolve merge conflicts
d6123ba
@seunghwak
fix build error
81f51c1
@seunghwak
update dataframe buffer utilities
69cb4f9
@seunghwak
reduce # resizes
6adcccb
@seunghwak
remove debug statement
bfe21fc
@seunghwak
rename VertexFrontierBucketType to KeyBucketType
446435b
@seunghwak
update per_v_transform_reduce_incoming|outgoing_e to support reduce_o…
df463ce 
…p::any
@seunghwak
update kernels to take KeyIterator key_first & key_last
222148d
@seunghwak
update per_v_transform_reduce_incoming_outgoing_e to support key list
effc69c
@seunghwak
remove pred_op.cuh
d537290
@seunghwak
update per_v_transform_reduce_incoming|outgoing_e to take a predicate
cf92885
@seunghwak
Merge branch 'branch-24.10' of https://github.com/rapidsai/cugraph in…
b8d846c 
…to enh_bfs_mg
@seunghwak
Merge branch 'branch-24.10' of https://github.com/rapidsai/cugraph in…
50ffc33 
…to enh_bfs_mg
@seunghwak
split per_v_transform_reduce_incoming_outgoing_e implementation to tw…
a6476b9 
…o files
@seunghwak
implement per_v_transform_reduce_if_incoming|outgoing_e
ec24758
@seunghwak
update BFS to use per_v_transform_reduce_if_outoging_e
df751e7
@seunghwak
file rename
4661b9b
@seunghwak
remove transform_reduce_v_frontier_outgoing_e_by_src (this can be bet…
0951741 
…ter supported by per_v_transform_reduce_if_outgoing_e)
@seunghwak
Merge branch 'branch-24.10' of https://github.com/rapidsai/cugraph in…
796b928 
…to enh_bfs_mg
@seunghwak
code cleanup, add few FIXMEs to improve performance, and add performa…
7b98e3a 
…nce measurement code
@seunghwak
performance tuning for BFS
3f77ee1
@seunghwak
add a utility to find iteator type in dataframe buffer
bb75771
@seunghwak
update unrenumber functions
4687949
@seunghwak
merge multiple allreduce opreations to a single allreduce operation
6c91fba
@seunghwak
update detail::per_v_transform_reudce_e
0bd734f
@seunghwak
added temporary code to configure NCCL to SHARP accelerate minor_comm
f1fb13c
@seunghwak
reduce printouts
f6d2b25
@seunghwak
bug fix in per_v_transform_reduce_e
ecb0c8e
@seunghwak
update detail::extract_transform_v_frontier_e
b419afb
@seunghwak
fine-tune direction optimizing alpha based on average vertex degree
7e718b1
@seunghwak
performance otimize local computing part of fill_edge_minor_property
58f3ac6
@seunghwak
parameter tuning in graph creation
b2dca9b
@seunghwak
reduce kernel launches in fill_edge_src_minor_property
ea0907d
@seunghwak
fix a build error
8a78131
@seunghwak
minor performance tuning
aa13925
@seunghwak
reduce comm. sync
2db13e9
@seunghwak
kernel fusion
a51b708
@seunghwak
kernel fusion
1526dcf
@seunghwak
fix regarding concurrent streams
a236533
@seunghwak
tune direction optimizing alpha
2ab3232
@seunghwak
update comments
65e36e1
@seunghwak
perf optimize direction optimizing aux info routines
d690435
@seunghwak
modify multi-stream kernel launching orders
8ecc412
@seunghwak
perf. opt. compute_vertex_list_bitmap_info
34c243c
@seunghwak
minor refactor
afa53fc
@seunghwak
additional performance optimizations
6265f13
@seunghwak
Graph 500 benchmark specific parameter tuning
6961fcd
@seunghwak
misc updates
f570034
@seunghwak
Graph 500 benchmark driver
c22f95c
@seunghwak
build only necessary files
c96c72b
@seunghwak
Copy link
Contributor Author

This PR is used to generate Graph 500 results and closing this PR for archival purpose.

Updates in this PR are pulled in PR #4751 and will be merged into the cugraph codebase using PR #4751.

@seunghwak seunghwak closed this Nov 12, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
CMake cuGraph
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
1 participant
@seunghwak