Skip to content

[Perf] Linux/x64: 14 Improvements on 5/5/2023 3:18:41 PM #17599

@performanceautofiler

Description

@performanceautofiler

Run Information

Name Value
Architecture x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Single>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 23.21 ns 2.76 ns 0.12 0.58 False
SquareRootBenchmark - Duration of single invocation 24.77 ns 2.72 ns 0.11 0.65 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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;Single&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Single>.NegateBenchmark


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 2.763560312521938 < 20.10756281804626.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 30.49048852735657 (T) = (0 -2.4321089117415724) / Math.Sqrt((142.35661605160854 / (299)) + (0.9239658437347126 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.9036836618918764 = (25.251260165345123 - 2.4321089117415724) / 25.251260165345123 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Single>.SquareRootBenchmark


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 2.7246690703251826 < 26.760846079818656.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 65.16745546267175 (T) = (0 -3.264226475557673) / Math.Sqrt((27.544430870408043 / (299)) + (0.8880779409241946 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.8924764040174462 = (30.358233890236615 - 3.264226475557673) / 30.358233890236615 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Int16>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 36.32 ns 2.31 ns 0.06 0.46 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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;Int16&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Int16>.NegateBenchmark


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 2.3109908674795676 < 33.217581843854994.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 8:38:27 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 86.29558520481004 (T) = (0 -3.045211053847624) / Math.Sqrt((21.362955087588478 / (299)) + (1.1713843248218914 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.9228023999088593 = (39.446965323434874 - 3.045211053847624) / 39.446965323434874 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<UInt16>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 35.54 ns 3.33 ns 0.09 0.43 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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;UInt16&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<UInt16>.NegateBenchmark


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 3.3287845114129655 < 32.95593447879347.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 67.25403596039794 (T) = (0 -3.0733422628513636) / Math.Sqrt((22.09063027962096 / (299)) + (2.4008739346370054 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.9220445042995304 = (39.4243181348002 - 3.0733422628513636) / 39.4243181348002 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Int

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 20.56 ns 1.57 ns 0.08 0.49 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.Runtime.Intrinsics.Tests.Perf_Vector128Int*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Int.NegateBenchmark


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 1.572498453547021 < 19.257490022263504.
IsChangePoint: Marked as a change because one of 4/12/2023 9:47:42 AM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 37.76802963551959 (T) = (0 -3.127234587632961) / Math.Sqrt((3.984586627798995 / (299)) + (2.783418124395193 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.861746103371488 = (22.61950414342272 - 3.127234587632961) / 22.61950414342272 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Double>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
SquareRootBenchmark - Duration of single invocation 18.26 ns 3.76 ns 0.21 0.70 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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;Double&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Double>.SquareRootBenchmark


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 3.758853222432616 < 17.8396394860596.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 43.941309170860514 (T) = (0 -3.039703586620027) / Math.Sqrt((20.789372528394157 / (299)) + (1.4446965210102312 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.8662887352736504 = (22.73333957943644 - 3.039703586620027) / 22.73333957943644 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Float

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
SquareRootBenchmark - Duration of single invocation 25.00 ns 2.22 ns 0.09 0.64 False
NegateBenchmark - Duration of single invocation 20.01 ns 3.05 ns 0.15 0.60 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.Runtime.Intrinsics.Tests.Perf_Vector128Float*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Float.SquareRootBenchmark


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 2.2205563521734377 < 25.44927971672359.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 43.94386370706255 (T) = (0 -2.9141703348519816) / Math.Sqrt((29.273681885482553 / (299)) + (3.1016223466747563 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.9028557669203747 = (29.998387371727468 - 2.9141703348519816) / 29.998387371727468 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Runtime.Intrinsics.Tests.Perf_Vector128Float.NegateBenchmark


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 3.048154284223983 < 20.00776381150514.
IsChangePoint: Marked as a change because one of 3/1/2023 5:53:50 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 46.77858391969227 (T) = (0 -3.4441813338133) / Math.Sqrt((3.8315567661734486 / (299)) + (1.7329591922057461 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.848618645169437 = (22.75168786584238 - 3.4441813338133) / 22.75168786584238 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Byte>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 85.80 ns 2.58 ns 0.03 0.26 True

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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;Byte&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Byte>.NegateBenchmark


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 2.581107232000826 < 81.76141564517626.
IsChangePoint: Marked as a change because one of 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 258.1340174894339 (T) = (0 -2.7838651226558784) / Math.Sqrt((29.125554546338 / (299)) + (0.10983591196211287 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.9681397597401998 = (87.37740519077126 - 2.7838651226558784) / 87.37740519077126 is greater than 0.05.
IsChangeEdgeDetector: Marked as 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Int32>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 20.14 ns 2.86 ns 0.14 0.50 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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;Int32&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Int32>.NegateBenchmark


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 2.858468655793391 < 19.473576319044955.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 84.85607895361731 (T) = (0 -2.6411385586131684) / Math.Sqrt((17.3504897792085 / (299)) + (0.02423909805712473 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.8874515720720825 = (23.46668547253902 - 2.6411385586131684) / 23.46668547253902 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<UInt32>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 22.39 ns 3.29 ns 0.15 0.59 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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;UInt32&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<UInt32>.NegateBenchmark


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 3.2934100154862196 < 19.67496865671986.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 65.65472637474534 (T) = (0 -2.7887691382723543) / Math.Sqrt((19.01325622581492 / (299)) + (0.42642791462418905 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.8827939606499273 = (23.793732419733228 - 2.7887691382723543) / 23.793732419733228 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<SByte>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 87.75 ns 1.05 ns 0.01 0.23 True

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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;SByte&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<SByte>.NegateBenchmark


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 1.0525644482921555 < 82.17746990648793.
IsChangePoint: Marked as a change because one of 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 200.9737234675778 (T) = (0 -2.971241396108933) / Math.Sqrt((28.067785462133635 / (299)) + (0.9041640620571089 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.9659655346150325 = (87.30095691237997 - 2.971241396108933) / 87.30095691237997 is greater than 0.05.
IsChangeEdgeDetector: Marked as 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 x64
OS ubuntu 18.04
Queue TigerUbuntu
Baseline b7c4ad80a32cdbcd2f87c20e5c7368c4043b02e4
Compare 640df60d36910d86691b9fbdb0aadb7724fa8493
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Int64>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
NegateBenchmark - Duration of single invocation 15.70 ns 3.18 ns 0.20 0.62 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.Runtime.Intrinsics.Tests.Perf_Vector128Of&lt;Int64&gt;*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<Int64>.NegateBenchmark


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 3.1762029887355965 < 12.601656224236981.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 5/4/2023 10:13:35 PM, 5/8/2023 3:52:55 PM falls between 4/29/2023 8:32:21 PM and 5/8/2023 3:52:55 PM.
IsImprovementStdDev: Marked as improvement because 32.686994740468165 (T) = (0 -3.250964431283636) / Math.Sqrt((20.610396533502826 / (299)) + (0.8071197292066078 / (11))) is greater than 1.9676960046166507 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (11) - 2, .975) and 0.7913589619959944 = (15.581615498007745 - 3.250964431283636) / 15.581615498007745 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

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions