Skip to content

Conversation

@christiangnrd
Copy link
Member

Fixes breakage from JuliaLang/julia#59165

@codecov

This comment was marked as resolved.

Copy link
Contributor

@github-actions github-actions bot left a comment

Choose a reason for hiding this comment

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

Metal Benchmarks

Benchmark suite Current: f14991d Previous: b6062f3 Ratio
latency/precompile 12279539458 ns 12123716500 ns 1.01
latency/ttfp 7980074500.5 ns 7891416041.5 ns 1.01
latency/import 1202277375 ns 1182082416.5 ns 1.02
integration/metaldevrt 942667 ns 942875 ns 1.00
integration/byval/slices=1 1674770.5 ns 1646750 ns 1.02
integration/byval/slices=3 20369875 ns 8867812.5 ns 2.30
integration/byval/reference 1669916 ns 1636542 ns 1.02
integration/byval/slices=2 2822125 ns 2692875 ns 1.05
kernel/indexing 500291.5 ns 695645.5 ns 0.72
kernel/indexing_checked 548896 ns 693625 ns 0.79
kernel/launch 14083 ns 13833 ns 1.02
array/construct 6167 ns 6250 ns 0.99
array/broadcast 620500 ns 673042 ns 0.92
array/random/randn/Float32 924583 ns 860292 ns 1.07
array/random/randn!/Float32 595916.5 ns 631250 ns 0.94
array/random/rand!/Int64 550125 ns 567500 ns 0.97
array/random/rand!/Float32 550791 ns 599083 ns 0.92
array/random/rand/Int64 902125 ns 774750 ns 1.16
array/random/rand/Float32 832958.5 ns 591188 ns 1.41
array/accumulate/Int64/1d 1377375 ns 1319666 ns 1.04
array/accumulate/Int64/dims=1 1944167 ns 1828042 ns 1.06
array/accumulate/Int64/dims=2 2361771 ns 2257542 ns 1.05
array/accumulate/Int64/dims=1L 12249396 ns 11786312.5 ns 1.04
array/accumulate/Int64/dims=2L 10109458.5 ns 9805667 ns 1.03
array/accumulate/Float32/1d 1160583.5 ns 1253250 ns 0.93
array/accumulate/Float32/dims=1 1693479.5 ns 1644041.5 ns 1.03
array/accumulate/Float32/dims=2 2137250 ns 1948625 ns 1.10
array/accumulate/Float32/dims=1L 10422584 ns 9938250 ns 1.05
array/accumulate/Float32/dims=2L 7586375 ns 7338500 ns 1.03
array/reductions/reduce/Int64/1d 1277292 ns 1363708.5 ns 0.94
array/reductions/reduce/Int64/dims=1 1201854 ns 1151250 ns 1.04
array/reductions/reduce/Int64/dims=2 1314209 ns 1290917 ns 1.02
array/reductions/reduce/Int64/dims=1L 2095292 ns 2102417 ns 1.00
array/reductions/reduce/Int64/dims=2L 3599541 ns 3586542 ns 1.00
array/reductions/reduce/Float32/1d 772875.5 ns 1074833 ns 0.72
array/reductions/reduce/Float32/dims=1 882729 ns 890729.5 ns 0.99
array/reductions/reduce/Float32/dims=2 753959 ns 804750 ns 0.94
array/reductions/reduce/Float32/dims=1L 1449292 ns 1386125 ns 1.05
array/reductions/reduce/Float32/dims=2L 1928875 ns 1922834 ns 1.00
array/reductions/mapreduce/Int64/1d 1245896 ns 1373771 ns 0.91
array/reductions/mapreduce/Int64/dims=1 1186500 ns 1156416 ns 1.03
array/reductions/mapreduce/Int64/dims=2 1306146 ns 1290792 ns 1.01
array/reductions/mapreduce/Int64/dims=1L 2036646 ns 2124875 ns 0.96
array/reductions/mapreduce/Int64/dims=2L 3519125 ns 3496375 ns 1.01
array/reductions/mapreduce/Float32/1d 811875 ns 1044417 ns 0.78
array/reductions/mapreduce/Float32/dims=1 872416 ns 896187.5 ns 0.97
array/reductions/mapreduce/Float32/dims=2 756458 ns 824145.5 ns 0.92
array/reductions/mapreduce/Float32/dims=1L 1416584 ns 1422916 ns 1.00
array/reductions/mapreduce/Float32/dims=2L 1931750 ns 1929583 ns 1.00
array/private/copyto!/gpu_to_gpu 556000 ns 670208 ns 0.83
array/private/copyto!/cpu_to_gpu 767145.5 ns 813458 ns 0.94
array/private/copyto!/gpu_to_cpu 679208.5 ns 828187.5 ns 0.82
array/private/iteration/findall/int 1685709 ns 1681459 ns 1.00
array/private/iteration/findall/bool 1529938 ns 1469917 ns 1.04
array/private/iteration/findfirst/int 1820917 ns 2055375 ns 0.89
array/private/iteration/findfirst/bool 1747166.5 ns 1844125 ns 0.95
array/private/iteration/scalar 2854083 ns 4211312 ns 0.68
array/private/iteration/logical 2798625 ns 2576333.5 ns 1.09
array/private/iteration/findmin/1d 1860542 ns 2090500 ns 0.89
array/private/iteration/findmin/2d 1666729 ns 1652354.5 ns 1.01
array/private/copy 823125 ns 586729.5 ns 1.40
array/shared/copyto!/gpu_to_gpu 82709 ns 78917 ns 1.05
array/shared/copyto!/cpu_to_gpu 80084 ns 82750 ns 0.97
array/shared/copyto!/gpu_to_cpu 79458 ns 81917 ns 0.97
array/shared/iteration/findall/int 1678645.5 ns 1683792 ns 1.00
array/shared/iteration/findall/bool 1556000 ns 1493979 ns 1.04
array/shared/iteration/findfirst/int 1476042 ns 1464375 ns 1.01
array/shared/iteration/findfirst/bool 1426250 ns 1449250 ns 0.98
array/shared/iteration/scalar 161583 ns 156250 ns 1.03
array/shared/iteration/logical 2354333 ns 2388291 ns 0.99
array/shared/iteration/findmin/1d 1550791.5 ns 1554062.5 ns 1.00
array/shared/iteration/findmin/2d 1666104 ns 1653667 ns 1.01
array/shared/copy 210042 ns 251666 ns 0.83
array/permutedims/4d 2593041 ns 2475958 ns 1.05
array/permutedims/2d 1273250 ns 1239375 ns 1.03
array/permutedims/3d 1880542 ns 1763541 ns 1.07
metal/synchronization/stream 15125 ns 14500 ns 1.04
metal/synchronization/context 16000 ns 15250 ns 1.05

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

@christiangnrd christiangnrd merged commit 3c4e045 into JuliaGPU:main Oct 14, 2025
16 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants