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] Linux/arm64: 3 Improvements on 5/10/2023 9:01:12 PM #17888

Closed
performanceautofiler bot opened this issue May 18, 2023 · 1 comment
Closed

[Perf] Linux/arm64: 3 Improvements on 5/10/2023 9:01:12 PM #17888

performanceautofiler bot opened this issue May 18, 2023 · 1 comment
Assignees
Labels
ampere arch-arm64 branch-refs/heads/main kind-micro os-linux perf-improvement PGO Applied if there were any profile guided optimization updates in the observed interval. runtime-coreclr untriaged

Comments

@performanceautofiler
Copy link

performanceautofiler bot commented May 18, 2023

Run Information

Name Value
Architecture arm64
OS ubuntu 20.04
Queue AmpereUbuntu
Baseline eead6a328cf3a7426e62d1441343b4b0e1f57d19
Compare 0b5f137ab0eea82f59d22da94a0b21f0bc598f16
Diff Diff
Configs CompilationMode:tiered, RunKind:micro

Improvements in System.IO.Compression.Brotli

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Decompress - Duration of single invocation 469.10 μs 426.36 μs 0.91 0.01 False

graph
Test Report

Repro

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

Payloads

Baseline
Compare

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.IO.Compression.Brotli*'

Payloads

Baseline
Compare

Histogram

System.IO.Compression.Brotli.Decompress(level: Fastest, file: "TestDocument.pdf")


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 426.35824794884167 < 445.8530575178268.
IsChangePoint: Marked as a change because one of 3/21/2023 2:53:21 PM, 4/12/2023 2:59:13 PM, 5/10/2023 2:36:50 PM, 5/18/2023 3:09:06 AM falls between 5/8/2023 3:52:55 PM and 5/18/2023 3:09:06 AM.
IsImprovementStdDev: Marked as improvement because 25.143119321571216 (T) = (0 -429792.5723166957) / Math.Sqrt((60320971.06795873 / (299)) + (26814158.968158115 / (19))) is greater than 1.9674995188240387 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (19) - 2, .975) and 0.06915990496612237 = (461725.4613436624 - 429792.5723166957) / 461725.4613436624 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

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


Run Information

Name Value
Architecture arm64
OS ubuntu 20.04
Queue AmpereUbuntu
Baseline eead6a328cf3a7426e62d1441343b4b0e1f57d19
Compare 0b5f137ab0eea82f59d22da94a0b21f0bc598f16
Diff Diff
Configs CompilationMode:tiered, RunKind:micro

Improvements in System.Reflection.Attributes

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
GetCustomAttributesMethodBaseMiss - Duration of single invocation 184.19 ns 158.62 ns 0.86 0.13 False
IsDefinedClassMissInherit - Duration of single invocation 1.15 μs 1.04 μs 0.90 0.05 False

graph
graph
Test Report

Repro

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

Payloads

Baseline
Compare

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Reflection.Attributes*'

Payloads

Baseline
Compare

Histogram

System.Reflection.Attributes.GetCustomAttributesMethodBaseMiss


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 158.6188600515804 < 175.60052049836727.
IsChangePoint: Marked as a change because one of 4/25/2023 6:51:20 AM, 5/10/2023 2:36:50 PM, 5/18/2023 3:09:06 AM falls between 5/8/2023 3:52:55 PM and 5/18/2023 3:09:06 AM.
IsImprovementStdDev: Marked as improvement because 17.48768447942659 (T) = (0 -159.98038967094453) / Math.Sqrt((35.342411653087545 / (299)) + (18.43658489537122 / (19))) is greater than 1.9674995188240387 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (19) - 2, .975) and 0.10237298970581 = (178.22590879758872 - 159.98038967094453) / 178.22590879758872 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Reflection.Attributes.IsDefinedClassMissInherit


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 1.0405882097915948 < 1.096303749249695.
IsChangePoint: Marked as a change because one of 5/10/2023 2:36:50 PM, 5/18/2023 3:09:06 AM falls between 5/8/2023 3:52:55 PM and 5/18/2023 3:09:06 AM.
IsImprovementStdDev: Marked as improvement because 23.71096365955759 (T) = (0 -991.4234120952227) / Math.Sqrt((992.883899451855 / (299)) + (926.3050943980411 / (19))) is greater than 1.9674995188240387 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (19) - 2, .975) and 0.1471821206561724 = (1162.5265324620545 - 991.4234120952227) / 1162.5265324620545 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

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

@kunalspathak
Copy link
Member

mostly from dotnet/runtime#86014

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
ampere arch-arm64 branch-refs/heads/main kind-micro os-linux perf-improvement PGO Applied if there were any profile guided optimization updates in the observed interval. runtime-coreclr untriaged
Projects
None yet
Development

No branches or pull requests

2 participants