Skip to content

Remove kron methods and use those in GPUArrays - #2643

Merged
maleadt merged 1 commit into
masterfrom
ksh/nokron
Feb 6, 2025
Merged

maleadt merged 1 commit into
masterfrom
ksh/nokron

Conversation

@kshyatt

@kshyatt kshyatt commented Feb 5, 2025

Copy link
Copy Markdown
Member

Should close #2602

@kshyatt kshyatt added the cuda array Stuff about CuArray. label Feb 5, 2025
@kshyatt
kshyatt requested a review from maleadt February 5, 2025 22:31
@codecov

codecov Bot commented Feb 6, 2025 •

Copy link
Copy Markdown

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 73.52%. Comparing base (f3b3f8b) to head (04bb133).
Report is 2 commits behind head on master.

Additional details and impacted files
@@            Coverage Diff             @@
##           master    #2643      +/-   ##
==========================================
+ Coverage   73.33%   73.52%   +0.19%     
==========================================
  Files         157      157              
  Lines       15306    15246      -60     
==========================================
- Hits        11225    11210      -15     
+ Misses       4081     4036      -45     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@github-actions github-actions Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

CUDA.jl Benchmarks

Details
Benchmark suite Current: 04bb133 Previous: f3b3f8b Ratio
latency/precompile 46646365401 ns 46487736314.5 ns 1.00
latency/ttfp 6955073219 ns 6967429207 ns 1.00
latency/import 3636121962 ns 3635420430 ns 1.00
integration/volumerhs 9626071 ns 9616119 ns 1.00
integration/byval/slices=1 146674 ns 146828 ns 1.00
integration/byval/slices=3 425077 ns 425349 ns 1.00
integration/byval/reference 144795 ns 145094 ns 1.00
integration/byval/slices=2 286075.5 ns 286178 ns 1.00
integration/cudadevrt 103348 ns 103633 ns 1.00
kernel/indexing 13884 ns 14176 ns 0.98
kernel/indexing_checked 14603 ns 14982 ns 0.97
kernel/occupancy 640.8047337278107 ns 632.635294117647 ns 1.01
kernel/launch 2023 ns 2070.8 ns 0.98
kernel/rand 14475 ns 14886 ns 0.97
array/reverse/1d 19560 ns 19934 ns 0.98
array/reverse/2d 24721 ns 25204 ns 0.98
array/reverse/1d_inplace 11370 ns 10780 ns 1.05
array/reverse/2d_inplace 12912 ns 12152 ns 1.06
array/copy 20976 ns 21364 ns 0.98
array/iteration/findall/int 155644 ns 152326.5 ns 1.02
array/iteration/findall/bool 133833.5 ns 137240.5 ns 0.98
array/iteration/findfirst/int 147283.5 ns 150634 ns 0.98
array/iteration/findfirst/bool 153190 ns 154806 ns 0.99
array/iteration/scalar 61915 ns 61510 ns 1.01
array/iteration/logical 204080.5 ns 208844.5 ns 0.98
array/iteration/findmin/1d 38824 ns 40358 ns 0.96
array/iteration/findmin/2d 93499.5 ns 94387 ns 0.99
array/reductions/reduce/1d 38402.5 ns 40907 ns 0.94
array/reductions/reduce/2d 46510.5 ns 52058 ns 0.89
array/reductions/mapreduce/1d 29781 ns 40613 ns 0.73
array/reductions/mapreduce/2d 48587 ns 52268.5 ns 0.93
array/broadcast 20793 ns 21343 ns 0.97
array/copyto!/gpu_to_gpu 13448 ns 11974 ns 1.12
array/copyto!/cpu_to_gpu 207835 ns 211036 ns 0.98
array/copyto!/gpu_to_cpu 242538 ns 243920 ns 0.99
array/accumulate/1d 108134 ns 109433 ns 0.99
array/accumulate/2d 79896 ns 80154 ns 1.00
array/construct 1279.35 ns 1298.45 ns 0.99
array/random/randn/Float32 44153 ns 45214.5 ns 0.98
array/random/randn!/Float32 26359 ns 27040 ns 0.97
array/random/rand!/Int64 27043 ns 27226 ns 0.99
array/random/rand!/Float32 8649.666666666666 ns 8945 ns 0.97
array/random/rand/Int64 29702 ns 30137 ns 0.99
array/random/rand/Float32 13018 ns 13528 ns 0.96
array/permutedims/4d 61058.5 ns 62138 ns 0.98
array/permutedims/2d 55549 ns 55960 ns 0.99
array/permutedims/3d 56209 ns 56875 ns 0.99
array/sorting/1d 2775964 ns 2778928.5 ns 1.00
array/sorting/by 3367476.5 ns 3370074 ns 1.00
array/sorting/2d 1084924 ns 1085961 ns 1.00
cuda/synchronization/stream/auto 1022.4545454545455 ns 1045.8 ns 0.98
cuda/synchronization/stream/nonblocking 6288 ns 6394 ns 0.98
cuda/synchronization/stream/blocking 809.0297029702971 ns 814.6046511627907 ns 0.99
cuda/synchronization/context/auto 1169.7 ns 1171.5 ns 1.00
cuda/synchronization/context/nonblocking 6562.6 ns 6630.8 ns 0.99
cuda/synchronization/context/blocking 907.439024390244 ns 899.4418604651163 ns 1.01

This comment was automatically generated by workflow using github-action-benchmark.

@maleadt
maleadt merged commit 031d7b9 into master Feb 6, 2025
@maleadt
maleadt deleted the ksh/nokron branch February 6, 2025 05:59
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

cuda array Stuff about CuArray.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Remove kron functionality

2 participants