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.

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

[Perf] Windows/x64: 9 Improvements on 7/12/2023 9:48:15 PM #19881

Closed
performanceautofiler bot opened this issue Jul 18, 2023 · 0 comments
Closed

[Perf] Windows/x64: 9 Improvements on 7/12/2023 9:48:15 PM #19881

performanceautofiler bot opened this issue Jul 18, 2023 · 0 comments
Labels
arch-x64 branch-refs/heads/main kind-micro os-windows perf-improvement PGO Applied if there were any profile guided optimization updates in the observed interval. runtime-coreclr

Comments

@performanceautofiler
Copy link

performanceautofiler bot commented Jul 18, 2023

Run Information

Name Value
Architecture x64
OS Windows 10.0.18362
Queue TigerWindows
Baseline d59af2cf097acb100ad6a9cba652be34be6f4a2e
Compare d9f6b98de38081b71fa9b5a463f302288546435d
Diff Diff
Configs CompilationMode:tiered, RunKind:micro

Improvements in Devirtualization.EqualityComparer

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
ValueTupleCompare - Duration of single invocation
📝 - Benchmark Source
3.45 ns 1.26 ns 0.36 0.23 False 43.69306792166419 17.763598416308596 0.4065541574731339)
ValueTupleCompareWrapped - Duration of single invocation
📝 - Benchmark Source
3.43 ns 0.93 ns 0.27 0.35 False 43.65031925696862 18.769477865624324 0.4299963479105102)
ValueTupleCompareCached - Duration of single invocation
📝 - Benchmark Source
4.42 ns 1.79 ns 0.41 0.30 False
ValueTupleCompareNoOpt - Duration of single invocation
📝 - Benchmark Source
7.09 ns 4.41 ns 0.62 0.44 False 68.14724427649156 45.93401803923917 0.6740407264727036)

graph
graph
graph
graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'Devirtualization.EqualityComparer*'

Payloads

Baseline
Compare

Devirtualization.EqualityComparer.ValueTupleCompare

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 1.2565320347846707 < 3.2822995026425765.
IsChangePoint: Marked as a change because one of 5/19/2023 10:30:38 AM, 6/16/2023 8:27:26 PM, 7/1/2023 6:36:20 PM, 7/5/2023 3:16:33 PM, 7/12/2023 4:12:15 PM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 24.730515224375335 (T) = (0 -0.8969217241015922) / Math.Sqrt((0.20408782872104278 / (299)) + (0.03104216550990351 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.6784598184992017 = (2.7894545556178505 - 0.8969217241015922) / 2.7894545556178505 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Devirtualization.EqualityComparer.ValueTupleCompareWrapped

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 0.9326349007457804 < 3.253045294606385.
IsChangePoint: Marked as a change because one of 7/1/2023 6:36:20 PM, 7/5/2023 3:16:33 PM, 7/12/2023 4:12:15 PM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 32.27263922197236 (T) = (0 -1.1998147831583863) / Math.Sqrt((0.15428070581587716 / (299)) + (0.017135526238660086 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.6096674625835075 = (3.0738272322867113 - 1.1998147831583863) / 3.0738272322867113 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Devirtualization.EqualityComparer.ValueTupleCompareCached

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 1.7941944875037978 < 4.198470651973903.
IsChangePoint: Marked as a change because one of 5/19/2023 10:30:38 AM, 7/5/2023 3:16:33 PM, 7/12/2023 7:15:00 AM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 62.61592073355686 (T) = (0 -1.7453224458052092) / Math.Sqrt((0.24413241619131412 / (299)) + (0.0005732691231184778 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.520031165187214 = (3.6363245261246604 - 1.7453224458052092) / 3.6363245261246604 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Devirtualization.EqualityComparer.ValueTupleCompareNoOpt

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 4.414920208090882 < 6.514658519753463.
IsChangePoint: Marked as a change because one of 7/5/2023 3:16:33 PM, 7/12/2023 4:12:15 PM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 3.947826219658748 (T) = (0 -4.530460493633428) / Math.Sqrt((0.519251974213766 / (299)) + (0.20961780808920732 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.14300938818784925 = (5.286476224113513 - 4.530460493633428) / 5.286476224113513 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository


Run Information

Name Value
Architecture x64
OS Windows 10.0.18362
Queue TigerWindows
Baseline d59af2cf097acb100ad6a9cba652be34be6f4a2e
Compare d9f6b98de38081b71fa9b5a463f302288546435d
Diff Diff
Configs CompilationMode:tiered, RunKind:micro

Improvements in BenchmarksGame.BinaryTrees_2

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
RunBench - Duration of single invocation
📝 - Benchmark Source
157.52 ms 135.51 ms 0.86 0.03 False 1189600000 1001775000 0.8421107935440484)

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'BenchmarksGame.BinaryTrees_2*'

Payloads

Baseline
Compare

BenchmarksGame.BinaryTrees_2.RunBench

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 135.5099575 < 150.1101329375.
IsChangePoint: Marked as a change because one of 5/19/2023 10:30:38 AM, 7/12/2023 4:12:15 PM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 20.250473266710692 (T) = (0 -131924217.25274725) / Math.Sqrt((71139015725084.39 / (299)) + (4293436764448.663 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.13035092335292267 = (151698220.34582 - 131924217.25274725) / 151698220.34582 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository


Run Information

Name Value
Architecture x64
OS Windows 10.0.18362
Queue TigerWindows
Baseline d59af2cf097acb100ad6a9cba652be34be6f4a2e
Compare d9f6b98de38081b71fa9b5a463f302288546435d
Diff Diff
Configs CompilationMode:tiered, RunKind:micro

Improvements in System.Collections.CreateAddAndClear<Int32>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
ImmutableList - Duration of single invocation
📝 - Benchmark Source
84.23 μs 76.45 μs 0.91 0.05 False 640857.1428571428 595529.5138888889 0.9292703070045079)

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Collections.CreateAddAndClear&lt;Int32&gt;*'

Payloads

Baseline
Compare

System.Collections.CreateAddAndClear<Int32>.ImmutableList(Size: 512)

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 76.44939746732027 < 80.31523164414412.
IsChangePoint: Marked as a change because one of 5/19/2023 10:30:38 AM, 7/12/2023 4:12:15 PM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 44.39911058696315 (T) = (0 -75766.02287985367) / Math.Sqrt((6230069.326437654 / (299)) + (184561.61087794177 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.11747308360739048 = (85851.23181234243 - 75766.02287985367) / 85851.23181234243 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository


Run Information

Name Value
Architecture x64
OS Windows 10.0.18362
Queue TigerWindows
Baseline d59af2cf097acb100ad6a9cba652be34be6f4a2e
Compare d9f6b98de38081b71fa9b5a463f302288546435d
Diff Diff
Configs CompilationMode:tiered, RunKind:micro

Improvements in System.Collections.ContainsTrue<Int32>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
ImmutableList - Duration of single invocation
📝 - Benchmark Source
635.33 μs 462.08 μs 0.73 0.05 False 6193965.517241379 5055808.080808081 0.8162473728235733)

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Collections.ContainsTrue&lt;Int32&gt;*'

Payloads

Baseline
Compare

System.Collections.ContainsTrue<Int32>.ImmutableList(Size: 512)

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 462.0780542986426 < 603.5381508333332.
IsChangePoint: Marked as a change because one of 5/22/2023 3:21:21 PM, 6/30/2023 8:34:57 AM, 7/12/2023 4:12:15 PM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 32.06492684580829 (T) = (0 -474236.7897135544) / Math.Sqrt((3701553056.9318767 / (299)) + (35686512.63219341 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.22448092535375225 = (611508.8657617829 - 474236.7897135544) / 611508.8657617829 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository


Run Information

Name Value
Architecture x64
OS Windows 10.0.18362
Queue TigerWindows
Baseline d59af2cf097acb100ad6a9cba652be34be6f4a2e
Compare d9f6b98de38081b71fa9b5a463f302288546435d
Diff Diff
Configs CompilationMode:tiered, RunKind:micro

Improvements in Benchstone.BenchI.Fib

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
Test - Duration of single invocation
📝 - Benchmark Source
159.72 μs 99.94 μs 0.63 0.04 False 1272411.0032362458 942972.6368159204 0.7410912310704378)

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'Benchstone.BenchI.Fib*'

Payloads

Baseline
Compare

Benchstone.BenchI.Fib.Test

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 99.93920259676628 < 151.74362288232604.
IsChangePoint: Marked as a change because one of 5/26/2023 6:22:05 AM, 6/13/2023 11:57:38 AM, 6/19/2023 6:37:24 PM, 6/30/2023 8:34:57 AM, 7/12/2023 4:12:15 PM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 22.301859234286507 (T) = (0 -104440.55985362288) / Math.Sqrt((56991740.20592351 / (299)) + (26577033.76880598 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.31459184812542557 = (152377.1778435674 - 104440.55985362288) / 152377.1778435674 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository


Run Information

Name Value
Architecture x64
OS Windows 10.0.18362
Queue TigerWindows
Baseline d59af2cf097acb100ad6a9cba652be34be6f4a2e
Compare d9f6b98de38081b71fa9b5a463f302288546435d
Diff Diff
Configs CompilationMode:tiered, RunKind:micro

Improvements in Benchstone.BenchI.Ackermann

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
Test - Duration of single invocation
📝 - Benchmark Source
3.53 μs 2.94 μs 0.83 0.06 False 23682.795698924732 20559.304314769055 0.8681113740174902)

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'Benchstone.BenchI.Ackermann*'

Payloads

Baseline
Compare

Benchstone.BenchI.Ackermann.Test

ETL Files

Baseline
Compare

Histogram

Description of detection logic

IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsRegressionBase: Marked as not a regression because the compare was not 5% greater than the baseline, or the value was too small.
IsImprovementBase: Marked as improvement because the compare was 5% less than the baseline, and the value was not too small.
IsImprovementCheck: Marked as improvement because the three check build points were 0.05 less than the baseline.
IsImprovementWindowed: Marked as improvement because 2.9422995364279902 < 3.3500482845055726.
IsChangePoint: Marked as a change because one of 4/25/2023 3:46:39 AM, 5/1/2023 3:42:23 PM, 5/9/2023 1:22:51 PM, 7/12/2023 4:12:15 PM, 7/17/2023 6:26:12 PM falls between 7/8/2023 10:33:35 PM and 7/17/2023 6:26:12 PM.
IsImprovementStdDev: Marked as improvement because 45.65593780180842 (T) = (0 -2938.4271900194885) / Math.Sqrt((21426.638209423287 / (299)) + (383.9760338270061 / (6))) is greater than 1.967824097752068 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (6) - 2, .975) and 0.15323750041750525 = (3470.1905096981873 - 2938.4271900194885) / 3470.1905096981873 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Baseline
Compare
Diff

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
arch-x64 branch-refs/heads/main kind-micro os-windows perf-improvement PGO Applied if there were any profile guided optimization updates in the observed interval. runtime-coreclr
Projects
None yet
Development

No branches or pull requests

1 participant