Bump UnsafeAtomicsLLVM to 0.2

[pxl-th]

Needed for AMDGPU backend to enable hardware atomics.

[github-actions]

Benchmark Results

	main	6e7c349...	main/6e7c349192bca3...
saxpy/default/Float16/1024	2.79 ± 0.19 μs	2.79 ± 0.19 μs	0.999
saxpy/default/Float16/1048576	2.07 ± 0.012 ms	2.07 ± 0.016 ms	0.998
saxpy/default/Float16/16384	0.0328 ± 0.00014 ms	0.0328 ± 0.00013 ms	1
saxpy/default/Float16/2048	5.21 ± 0.02 μs	5.22 ± 0.02 μs	0.999
saxpy/default/Float16/256	1.7 ± 0.024 μs	1.71 ± 0.027 μs	0.994
saxpy/default/Float16/262144	0.516 ± 0.0012 ms	0.516 ± 0.0015 ms	1
saxpy/default/Float16/32768	0.065 ± 0.00017 ms	0.065 ± 0.00018 ms	1
saxpy/default/Float16/4096	10.1 ± 0.061 μs	10.1 ± 0.06 μs	1
saxpy/default/Float16/512	2.08 ± 0.036 μs	2.08 ± 0.046 μs	1
saxpy/default/Float16/64	1.42 ± 0.009 μs	1.42 ± 0.01 μs	1.01
saxpy/default/Float16/65536	0.129 ± 0.00031 ms	0.129 ± 0.00035 ms	1
saxpy/default/Float32/1024	1.04 ± 0.17 μs	1.03 ± 0.16 μs	1.01
saxpy/default/Float32/1048576	0.885 ± 0.013 ms	0.886 ± 0.014 ms	0.999
saxpy/default/Float32/16384	14.4 ± 0.12 μs	14.4 ± 0.12 μs	1
saxpy/default/Float32/2048	1.73 ± 0.18 μs	1.73 ± 0.18 μs	1
saxpy/default/Float32/256	1.24 ± 0.014 μs	1.22 ± 0.012 μs	1.01
saxpy/default/Float32/262144	0.221 ± 0.00059 ms	0.221 ± 0.0006 ms	1
saxpy/default/Float32/32768	28.3 ± 0.12 μs	28.3 ± 0.14 μs	0.999
saxpy/default/Float32/4096	3.04 ± 0.02 μs	3.03 ± 0.023 μs	1
saxpy/default/Float32/512	1.28 ± 0.013 μs	1.26 ± 0.012 μs	1.02
saxpy/default/Float32/64	1.21 ± 0.012 μs	1.2 ± 0.012 μs	1.01
saxpy/default/Float32/65536	0.056 ± 0.00026 ms	0.0563 ± 0.00039 ms	0.995
saxpy/default/Float64/1024	1.07 ± 0.17 μs	1.07 ± 0.16 μs	0.998
saxpy/default/Float64/1048576	1.01 ± 0.028 ms	1.01 ± 0.031 ms	0.991
saxpy/default/Float64/16384	16.3 ± 0.6 μs	16.5 ± 0.41 μs	0.988
saxpy/default/Float64/2048	1.8 ± 0.16 μs	1.79 ± 0.16 μs	1
saxpy/default/Float64/256	1.34 ± 0.011 μs	1.35 ± 0.015 μs	0.992
saxpy/default/Float64/262144	0.243 ± 0.0043 ms	0.245 ± 0.0074 ms	0.994
saxpy/default/Float64/32768	0.0317 ± 0.00095 ms	0.0325 ± 0.0011 ms	0.976
saxpy/default/Float64/4096	3.06 ± 0.048 μs	3.07 ± 0.054 μs	0.997
saxpy/default/Float64/512	1.36 ± 0.015 μs	1.37 ± 0.017 μs	0.996
saxpy/default/Float64/64	1.34 ± 0.02 μs	1.34 ± 0.026 μs	0.997
saxpy/default/Float64/65536	0.063 ± 0.0013 ms	0.0621 ± 0.0019 ms	1.02
saxpy/static workgroup=(1024,)/Float16/1024	2.06 ± 0.2 μs	2.07 ± 0.21 μs	0.992
saxpy/static workgroup=(1024,)/Float16/1048576	0.163 ± 0.0072 ms	0.197 ± 0.021 ms	0.828
saxpy/static workgroup=(1024,)/Float16/16384	4.28 ± 0.21 μs	4.25 ± 0.21 μs	1.01
saxpy/static workgroup=(1024,)/Float16/2048	2.1 ± 0.21 μs	2.12 ± 0.21 μs	0.987
saxpy/static workgroup=(1024,)/Float16/256	2.61 ± 0.045 μs	2.63 ± 0.037 μs	0.992
saxpy/static workgroup=(1024,)/Float16/262144	0.044 ± 0.0024 ms	0.0484 ± 0.0046 ms	0.909
saxpy/static workgroup=(1024,)/Float16/32768	6.8 ± 0.26 μs	6.85 ± 0.25 μs	0.993
saxpy/static workgroup=(1024,)/Float16/4096	2.38 ± 0.03 μs	2.41 ± 0.033 μs	0.988
saxpy/static workgroup=(1024,)/Float16/512	3.13 ± 0.058 μs	3.14 ± 0.069 μs	0.996
saxpy/static workgroup=(1024,)/Float16/64	2.24 ± 0.023 μs	2.25 ± 0.022 μs	0.997
saxpy/static workgroup=(1024,)/Float16/65536	12.9 ± 0.53 μs	13.2 ± 1.2 μs	0.976
saxpy/static workgroup=(1024,)/Float32/1024	1.93 ± 0.023 μs	1.93 ± 0.023 μs	0.998
saxpy/static workgroup=(1024,)/Float32/1048576	0.247 ± 0.0056 ms	0.275 ± 0.032 ms	0.9
saxpy/static workgroup=(1024,)/Float32/16384	4.8 ± 0.28 μs	4.85 ± 0.23 μs	0.988
saxpy/static workgroup=(1024,)/Float32/2048	2.26 ± 0.22 μs	2.25 ± 0.22 μs	1
saxpy/static workgroup=(1024,)/Float32/256	2.84 ± 1.9 μs	2.76 ± 1.6 μs	1.03
saxpy/static workgroup=(1024,)/Float32/262144	0.0632 ± 0.0056 ms	0.0628 ± 0.0098 ms	1.01
saxpy/static workgroup=(1024,)/Float32/32768	8.26 ± 0.97 μs	8.26 ± 1 μs	0.999
saxpy/static workgroup=(1024,)/Float32/4096	2.42 ± 0.19 μs	2.4 ± 0.19 μs	1.01
saxpy/static workgroup=(1024,)/Float32/512	2.43 ± 0.22 μs	2.45 ± 0.21 μs	0.991
saxpy/static workgroup=(1024,)/Float32/64	2.41 ± 0.047 μs	2.42 ± 0.055 μs	0.996
saxpy/static workgroup=(1024,)/Float32/65536	17.7 ± 2.8 μs	18.2 ± 3.4 μs	0.976
saxpy/static workgroup=(1024,)/Float64/1024	2.05 ± 0.055 μs	2.04 ± 0.063 μs	1
saxpy/static workgroup=(1024,)/Float64/1048576	0.461 ± 0.061 ms	0.542 ± 0.071 ms	0.85
saxpy/static workgroup=(1024,)/Float64/16384	8.32 ± 0.83 μs	8.5 ± 0.85 μs	0.979
saxpy/static workgroup=(1024,)/Float64/2048	2.51 ± 0.3 μs	2.51 ± 0.24 μs	0.999
saxpy/static workgroup=(1024,)/Float64/256	2.39 ± 0.042 μs	2.41 ± 0.052 μs	0.993
saxpy/static workgroup=(1024,)/Float64/262144	0.117 ± 0.015 ms	0.117 ± 0.021 ms	0.993
saxpy/static workgroup=(1024,)/Float64/32768	17.3 ± 2.7 μs	16.7 ± 2.7 μs	1.04
saxpy/static workgroup=(1024,)/Float64/4096	3.02 ± 0.33 μs	3.02 ± 0.31 μs	1
saxpy/static workgroup=(1024,)/Float64/512	2.38 ± 0.038 μs	2.4 ± 0.042 μs	0.992
saxpy/static workgroup=(1024,)/Float64/64	2.37 ± 0.073 μs	2.39 ± 0.081 μs	0.992
saxpy/static workgroup=(1024,)/Float64/65536	0.0321 ± 0.0044 ms	31.5 ± 4.5 μs	1.02
time_to_load	0.454 ± 0.0021 s	0.466 ± 0.0024 s	0.975

Benchmark Plots

A plot of the benchmark results have been uploaded as an artifact to the workflow run for this PR.
Go to "Actions"->"Benchmark a pull request"->[the most recent run]->"Artifacts" (at the bottom).