[Perf] Linux/x64: 27 Improvements on 4/19/2023 8:38:27 PM

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	ubuntu 18.04
Queue	TigerUbuntu
Baseline	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
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
XorBenchmark - Duration of single invocation	5.01 ns	2.68 ns	0.53	0.66	False
OnesComplementOperatorBenchmark - Duration of single invocation	5.42 ns	4.24 ns	0.78	0.64	False

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<SByte>*'

Payloads

Baseline
Compare

Histogram

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

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 2.679743357250427 < 3.6449466262683385.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 6.820953320836807 (T) = (0 -3.018439216071591) / Math.Sqrt((88.83718814319592 / (299)) + (0.8352664617023011 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.5655761706271869 = (6.948143752678981 - 3.018439216071591) / 6.948143752678981 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

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

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 4.235926873612313 < 5.224453875999279.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 6.869257634031004 (T) = (0 -4.455212494375702) / Math.Sqrt((55.740725703833654 / (299)) + (1.26312107500523 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.4298282599942029 = (7.813807984117915 - 4.455212494375702) / 7.813807984117915 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tests.Perf_VectorOf<Int64>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
XorBenchmark - Duration of single invocation	4.35 ns	1.46 ns	0.34	0.59	False

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.Numerics.Tests.Perf_VectorOf<Int64>*'

Payloads

Baseline
Compare

Histogram

System.Numerics.Tests.Perf_VectorOf<Int64>.XorBenchmark

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 1.4614028928738456 < 3.8317845152241055.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 5.523866695759602 (T) = (0 -3.097826622144042) / Math.Sqrt((102.85248078646798 / (299)) + (1.1460704115305733 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.5274225424211153 = (6.555172220898706 - 3.097826622144042) / 6.555172220898706 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tests.Perf_VectorOf<Int16>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
XorBenchmark - Duration of single invocation	3.52 ns	2.52 ns	0.72	0.56	False
MaxBenchmark - Duration of single invocation	5.25 ns	1.46 ns	0.28	0.53	False

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.Numerics.Tests.Perf_VectorOf<Int16>*'

Payloads

Baseline
Compare

Histogram

System.Numerics.Tests.Perf_VectorOf<Int16>.XorBenchmark

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 2.516420384255149 < 3.1053436114479.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 6.162387678361167 (T) = (0 -2.897116699810145) / Math.Sqrt((100.97635738268015 / (299)) + (0.17562391443654285 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.5554431063921368 = (6.516863738852844 - 2.897116699810145) / 6.516863738852844 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Numerics.Tests.Perf_VectorOf<Int16>.MaxBenchmark

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 1.464161090598706 < 3.2230674999611097.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 12.074585760006482 (T) = (0 -3.3301357748658877) / Math.Sqrt((653.708985978887 / (299)) + (1.1138607225861885 / (23))) is greater than 1.9674049737480528 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (23) - 2, .975) and 0.8442443712373853 = (21.380516398166947 - 3.3301357748658877) / 21.380516398166947 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tests.Perf_VectorOf<Byte>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
OnesComplementOperatorBenchmark - Duration of single invocation	5.79 ns	2.68 ns	0.46	0.66	False
ExclusiveOrOperatorBenchmark - Duration of single invocation	4.34 ns	2.69 ns	0.62	0.59	False

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.Numerics.Tests.Perf_VectorOf<Byte>*'

Payloads

Baseline
Compare

Histogram

System.Numerics.Tests.Perf_VectorOf<Byte>.OnesComplementOperatorBenchmark

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 2.676323898732861 < 4.346276047984768.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 6.981877737635318 (T) = (0 -3.0136049970780676) / Math.Sqrt((51.58664715212985 / (299)) + (1.677859721007907 / (23))) is greater than 1.9674049737480528 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (23) - 2, .975) and 0.5344237128922814 = (6.472848984211306 - 3.0136049970780676) / 6.472848984211306 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Numerics.Tests.Perf_VectorOf<Byte>.ExclusiveOrOperatorBenchmark

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 2.6896873564109853 < 3.152707466143713.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 5.937705385290969 (T) = (0 -2.9679604692800585) / Math.Sqrt((102.72160996605471 / (299)) + (0.13282712782392103 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.541709802313618 = (6.47615961297758 - 2.9679604692800585) / 6.47615961297758 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tests.Perf_VectorOf<SByte>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
LessThanAnyBenchmark - Duration of single invocation	6.69 ns	0.00 ns	0.00	0.67	False
XorBenchmark - Duration of single invocation	3.43 ns	2.17 ns	0.63	0.66	False

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.Numerics.Tests.Perf_VectorOf<SByte>*'

Payloads

Baseline
Compare

Histogram

System.Numerics.Tests.Perf_VectorOf<SByte>.LessThanAnyBenchmark

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 < 2.3685088284722293.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 9.525922697693456 (T) = (0 -2.2820053611489444) / Math.Sqrt((27.024018869857915 / (299)) + (1.2748786341377607 / (23))) is greater than 1.9674049737480528 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (23) - 2, .975) and 0.6144933417903515 = (5.919496622307185 - 2.2820053611489444) / 5.919496622307185 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Numerics.Tests.Perf_VectorOf<SByte>.XorBenchmark

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 2.1695191129640543 < 3.2796983386831275.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 5.829763912188766 (T) = (0 -3.019303046626311) / Math.Sqrt((100.42349900078176 / (299)) + (0.7209355716424958 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.538736642979942 = (6.545724911105428 - 3.019303046626311) / 6.545724911105428 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tests.Perf_VectorOf<UInt32>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
DivisionOperatorBenchmark - Duration of single invocation	23.02 ns	21.75 ns	0.95	0.46	False
ExclusiveOrOperatorBenchmark - Duration of single invocation	4.56 ns	2.72 ns	0.60	0.60	False
DivideBenchmark - Duration of single invocation	29.97 ns	21.34 ns	0.71	0.48	False

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.Numerics.Tests.Perf_VectorOf<UInt32>*'

Payloads

Baseline
Compare

Histogram

System.Numerics.Tests.Perf_VectorOf<UInt32>.DivisionOperatorBenchmark

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 21.754411940313485 < 22.368488728133244.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 9.116050173549471 (T) = (0 -22.334510941838516) / Math.Sqrt((9.55829861582482 / (299)) + (2.040806151043863 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.12250899722071754 = (25.45269509442067 - 22.334510941838516) / 25.45269509442067 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Numerics.Tests.Perf_VectorOf<UInt32>.ExclusiveOrOperatorBenchmark

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 2.718198875212073 < 3.1678990102090197.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 4.5900635962686485 (T) = (0 -3.508004080664486) / Math.Sqrt((102.46254317834466 / (299)) + (2.321823468906372 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.4644860296773509 = (6.550723744052655 - 3.508004080664486) / 6.550723744052655 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Numerics.Tests.Perf_VectorOf<UInt32>.DivideBenchmark

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 21.337840326777293 < 22.387926520868486.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 7.86734636466586 (T) = (0 -22.387157359454985) / Math.Sqrt((17.74649484287698 / (299)) + (2.5289950664191143 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.12482647382509213 = (25.58024973321782 - 22.387157359454985) / 25.58024973321782 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
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
ConditionalSelectBenchmark - Duration of single invocation	4.12 ns	2.79 ns	0.68	0.64	False
EqualityOperatorBenchmark - Duration of single invocation	2.31 ns	0.00 ns	0.00	0.74	False
SubtractionOperatorBenchmark - Duration of single invocation	5.53 ns	2.72 ns	0.49	0.59	False

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<UInt32>*'

Payloads

Baseline
Compare

Histogram

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

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 2.7932632964485755 < 3.7028329606441286.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 7.284685320833955 (T) = (0 -3.387584608845858) / Math.Sqrt((97.86350730115872 / (299)) + (2.067441896327898 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.5803092531873341 = (8.071620912714446 - 3.387584608845858) / 8.071620912714446 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

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

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 < 2.226958903692777.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 10.587452603680207 (T) = (0 -2.1038231359763975) / Math.Sqrt((26.73712650829325 / (299)) + (0.6098861348928829 / (23))) is greater than 1.9674049737480528 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (23) - 2, .975) and 0.6314788055692351 = (5.708825347823105 - 2.1038231359763975) / 5.708825347823105 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

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

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 2.7179373106980496 < 3.1788043232979546.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 12.625477906084814 (T) = (0 -2.977179148104469) / Math.Sqrt((133.37840009077846 / (299)) + (0.29406781874744164 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.741669219630364 = (11.524678336218908 - 2.977179148104469) / 11.524678336218908 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tests.Perf_VectorOf<UInt64>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
GreaterThanOrEqualBenchmark - Duration of single invocation	6.38 ns	4.17 ns	0.65	0.60	False

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.Numerics.Tests.Perf_VectorOf<UInt64>*'

Payloads

Baseline
Compare

Histogram

System.Numerics.Tests.Perf_VectorOf<UInt64>.GreaterThanOrEqualBenchmark

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 4.171646186915987 < 4.391062665102214.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 12.192995157401569 (T) = (0 -3.9697798988198834) / Math.Sqrt((45.72438275334887 / (299)) + (0.3958798758962949 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.5583484074510433 = (8.98848768077257 - 3.9697798988198834) / 8.98848768077257 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
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
AndNotBenchmark - Duration of single invocation	4.34 ns	3.25 ns	0.75	0.63	False

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<UInt16>*'

Payloads

Baseline
Compare

Histogram

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

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 3.2522920988093755 < 3.359202191332246.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 6.493162471832812 (T) = (0 -3.124410408460843) / Math.Sqrt((87.68376352110583 / (299)) + (1.0063158333231923 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.546111128952389 = (6.8836462133329235 - 3.124410408460843) / 6.8836462133329235 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

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	ubuntu 18.04
Queue	TigerUbuntu
Baseline	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
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
ConditionalSelectBenchmark - Duration of single invocation	3.90 ns	2.71 ns	0.70	0.65	False
LessThanOrEqualBenchmark - Duration of single invocation	6.74 ns	2.97 ns	0.44	0.67	False

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<Single>*'

Payloads

Baseline
Compare

Histogram

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

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 2.7098771274110494 < 3.444335056357735.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 7.508657775497764 (T) = (0 -3.3491348044456006) / Math.Sqrt((102.89578760290418 / (299)) + (2.2833559839955244 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.5977352489830465 = (8.32569792898521 - 3.3491348044456006) / 8.32569792898521 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>.LessThanOrEqualBenchmark

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 2.970279442827728 < 3.2479824581826127.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 10.646920298230135 (T) = (0 -3.2634318071599893) / Math.Sqrt((199.13382436899911 / (299)) + (2.6339047151246433 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.7418346125138932 = (12.640857238601013 - 3.2634318071599893) / 12.640857238601013 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Runtime.Intrinsics.Tests.Perf_Vector128Of<UInt64>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
GreaterThanOrEqualBenchmark - Duration of single invocation	4.70 ns	3.46 ns	0.73	0.67	False
SubtractionOperatorBenchmark - Duration of single invocation	6.29 ns	2.66 ns	0.42	0.66	False

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<UInt64>*'

Payloads

Baseline
Compare

Histogram

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<UInt64>.GreaterThanOrEqualBenchmark

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 3.455636340914698 < 4.207782403568587.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 11.511597558284661 (T) = (0 -3.677997676383146) / Math.Sqrt((150.2051773778856 / (299)) + (1.493884610202136 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.7016507226536257 = (12.327824987868514 - 3.677997676383146) / 12.327824987868514 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Runtime.Intrinsics.Tests.Perf_Vector128Of<UInt64>.SubtractionOperatorBenchmark

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 2.6644641353066807 < 4.04579036812246.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 11.544430096359823 (T) = (0 -3.0894304332317133) / Math.Sqrt((52.73994638636583 / (299)) + (0.8623467851287484 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.6325991258975601 = (8.408881554185712 - 3.0894304332317133) / 8.408881554185712 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tests.Perf_VectorOf<UInt16>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
MultiplyOperatorBenchmark - Duration of single invocation	3.65 ns	2.17 ns	0.59	0.48	False

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.Numerics.Tests.Perf_VectorOf<UInt16>*'

Payloads

Baseline
Compare

Histogram

System.Numerics.Tests.Perf_VectorOf<UInt16>.MultiplyOperatorBenchmark

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 2.173202530167462 < 3.115499839936436.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 12.133628876530844 (T) = (0 -3.174309039049458) / Math.Sqrt((517.0865181020355 / (299)) + (1.1744318769109674 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.8359949967808815 = (19.35495245110559 - 3.174309039049458) / 19.35495245110559 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
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
OnesComplementBenchmark - Duration of single invocation	4.11 ns	2.95 ns	0.72	0.61	False

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<Double>*'

Payloads

Baseline
Compare

Histogram

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

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 2.947704184495585 < 3.2056795316269833.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 5.672852434505756 (T) = (0 -3.0396565581581796) / Math.Sqrt((77.92070979996524 / (299)) + (0.6207991959739921 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.49972214267774584 = (6.075936629352324 - 3.0396565581581796) / 6.075936629352324 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
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
MinBenchmark - Duration of single invocation	3.50 ns	1.73 ns	0.49	0.64	False

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<Int64>*'

Payloads

Baseline
Compare

Histogram

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

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 1.730847662796137 < 3.090836547937163.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 10.798150293692668 (T) = (0 -3.1523959071107615) / Math.Sqrt((100.99777079599842 / (299)) + (0.6830007523049385 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.6745822169074337 = (9.687226915358991 - 3.1523959071107615) / 9.687226915358991 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tests.Perf_VectorConvert

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio	Baseline ETL	Compare ETL
Convert_double_long - Duration of single invocation	15.32 μs	13.51 μs	0.88	0.46	False
Convert_float_int - Duration of single invocation	18.26 μs	13.79 μs	0.76	0.45	False

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.Numerics.Tests.Perf_VectorConvert*'

Payloads

Baseline
Compare

Histogram

System.Numerics.Tests.Perf_VectorConvert.Convert_double_long

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 13.510735969939448 < 14.449117325236262.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 10.404307305871184 (T) = (0 -13530.240065641025) / Math.Sqrt((10520282.528597418 / (299)) + (653630.3277333181 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.1611567764425047 = (16129.641017137747 - 13530.240065641025) / 16129.641017137747 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Numerics.Tests.Perf_VectorConvert.Convert_float_int

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 13.7932641384778 < 18.322456842592064.
IsChangePoint: Marked as a change because one of 3/1/2023 3:09:21 PM, 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 29.323619510905253 (T) = (0 -13932.944249772141) / Math.Sqrt((7353820.839940711 / (299)) + (555908.3577350235 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.3150376566309624 = (20341.18281779102 - 13932.944249772141) / 20341.18281779102 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	7b662a7e264552ea19a51a0135f2441bd5f141aa
Compare	46e2597cbb9cbf68cf935ded056597afb492111e
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
OnesComplementBenchmark - Duration of single invocation	3.89 ns	2.75 ns	0.71	0.64	False

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<Int16>*'

Payloads

Baseline
Compare

Histogram

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

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 2.753108823911185 < 3.0795469832570235.
IsChangePoint: Marked as a change because one of 4/12/2023 8:39:35 PM, 4/19/2023 3:30:01 PM, 4/25/2023 3:46:39 AM falls between 4/12/2023 8:39:35 PM and 4/25/2023 3:46:39 AM.
IsImprovementStdDev: Marked as improvement because 6.0997094838345545 (T) = (0 -3.040314120006212) / Math.Sqrt((68.38958062348108 / (299)) + (0.19736050418926473 / (24))) is greater than 1.967381707010986 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (24) - 2, .975) and 0.49408263956420656 = (6.009507397388593 - 3.040314120006212) / 6.009507397388593 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

[lewing]

looks like dotnet/runtime#84304 and dotnet/runtime#85048