Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[Perf] Changes at 7/5/2022 10:32:06 PM #6721

Closed
performanceautofiler bot opened this issue Jul 12, 2022 · 0 comments
Closed

[Perf] Changes at 7/5/2022 10:32:06 PM #6721

performanceautofiler bot opened this issue Jul 12, 2022 · 0 comments
Assignees
@tannergooding
Labels
arch-x64 branch-refs/heads/main kind-micro os-linux perf-improvement runtime-coreclr

Comments

@performanceautofiler
Copy link

performanceautofiler bot commented Jul 12, 2022 •
edited
Loading

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchF.MDSqMtx

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 930.53 ms 770.26 ms 0.83 0.04 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchF.MDSqMtx*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchF.MDSqMtx.Test

Description of detection logic

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 770.2648947692308 < 885.1524572446428.
IsChangePoint: Marked as a change because one of 6/6/2022 10:12:27 PM, 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 32.968357364196336 (T) = (0 -762239601.1260234) / Math.Sqrt((1984588261209949.8 / (299)) + (465130434643554.44 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.1512434488371719 = (898066235.9327027 - 762239601.1260234) / 898066235.9327027 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

### Run Information
Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchI.MDGeneralArray

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 16.63 ms 11.17 ms 0.67 0.03 False
Test2 - Duration of single invocation 16.57 ms 11.16 ms 0.67 0.02 False

graph
graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchI.MDGeneralArray*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchI.MDGeneralArray.Test

Description of detection logic

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 11.173597879166666 < 15.745436724068512.
IsChangePoint: Marked as a change because one of 6/14/2022 10:05:19 AM, 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 34.68017868912036 (T) = (0 -11048083.262695614) / Math.Sqrt((2624996997198.8794 / (299)) + (76785529424.15703 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.24324601434485607 = (14599306.342775278 - 11048083.262695614) / 14599306.342775278 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

```#### Benchstone.MDBenchI.MDGeneralArray.Test2

```log

Description of detection logic

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 11.159814841145835 < 15.660590579609377.
IsChangePoint: Marked as a change because one of 6/14/2022 10:05:19 AM, 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 34.915333879275316 (T) = (0 -10751162.077037025) / Math.Sqrt((2590281324396.07 / (299)) + (147570572407.2758 / (44))) is greater than 1.9669451293272056 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (44) - 2, .975) and 0.2625410069884124 = (14578657.496781103 - 10751162.077037025) / 14578657.496781103 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchI.MDMidpoint

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 556.00 ms 413.91 ms 0.74 0.01 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchI.MDMidpoint*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchI.MDMidpoint.Test

Description of detection logic

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 413.9137170714286 < 527.9502956349999.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 160.14970836247554 (T) = (0 -413893152.28634006) / Math.Sqrt((64429995116528.63 / (299)) + (27811119708367.203 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.2610431713948699 = (560104645.1761101 - 413893152.28634006) / 560104645.1761101 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchF.MDInvMt

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 5.97 ms 2.55 ms 0.43 0.06 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchF.MDInvMt*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchF.MDInvMt.Test

Description of detection logic

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 2.5481520705357146 < 5.671169423869621.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 273.83275849387024 (T) = (0 -2543588.855900561) / Math.Sqrt((42794490307.911385 / (299)) + (84033987.2352634 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.5645203390217909 = (5840890.135229161 - 2543588.855900561) / 5840890.135229161 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchI.MDArray2

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 1.31 secs 1.03 secs 0.78 0.00 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchI.MDArray2*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchI.MDArray2.Test

Description of detection logic

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 1.0266930618461538 < 1.242550278826923.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 440.348503206845 (T) = (0 -1025749608.3284578) / Math.Sqrt((2332543136787.7573 / (299)) + (18195856383547.684 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.21605731666833128 = (1308449750.393915 - 1025749608.3284578) / 1308449750.393915 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchI.MDPuzzle

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 550.24 ms 415.93 ms 0.76 0.00 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchI.MDPuzzle*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchI.MDPuzzle.Test

Description of detection logic

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 415.9348021428571 < 522.6139306895834.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 415.8859307031019 (T) = (0 -417048645.4050549) / Math.Sqrt((2380625742831.7383 / (299)) + (4203366618506.9224 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.24098618427737317 = (549461204.4815015 - 417048645.4050549) / 549461204.4815015 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchI.MDAddArray2

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 17.51 ms 11.99 ms 0.68 0.03 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchI.MDAddArray2*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchI.MDAddArray2.Test

Description of detection logic

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 11.99241771875 < 17.02533243732804.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 92.27283746740929 (T) = (0 -11773072.591718813) / Math.Sqrt((166012199423.00952 / (299)) + (180987961031.2528 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.34651323094103886 = (18015778.052664112 - 11773072.591718813) / 18015778.052664112 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchI.MDXposMatrix

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 44.59 μs 33.72 μs 0.76 0.05 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchI.MDXposMatrix*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchI.MDXposMatrix.Test

Description of detection logic

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 33.715246545951345 < 42.357845826338014.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 42.27667062710874 (T) = (0 -32693.206483787755) / Math.Sqrt((31161.21119668241 / (299)) + (3620264.3331391825 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.26848114327597233 = (44692.22656842812 - 32693.206483787755) / 44692.22656842812 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in ByteMark

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
BenchNumericSortRectangular - Duration of single invocation 1.12 secs 861.35 ms 0.77 0.00 False
BenchLUDecomp - Duration of single invocation 1.01 secs 916.23 ms 0.91 0.04 False

graph
graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'ByteMark*'

Payloads

Baseline
Compare

Histogram

ByteMark.BenchNumericSortRectangular

Description of detection logic

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 861.3462037857142 < 1.0682378820841345.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 178.22353945623888 (T) = (0 -865166731.555458) / Math.Sqrt((116021972679181.58 / (299)) + (73076325628214.98 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.22612349537013468 = (1117964851.4710438 - 865166731.555458) / 1117964851.4710438 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

```#### ByteMark.BenchLUDecomp

```log

Description of detection logic

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 916.2305363333334 < 0.9526970249516666.
IsChangePoint: Marked as a change because one of 5/11/2022 7:02:26 PM, 5/31/2022 6:06:22 PM, 6/2/2022 5:58:32 PM, 6/7/2022 6:20:49 PM, 6/13/2022 2:27:15 PM, 7/10/2022 2:23:14 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 31.13495311479163 (T) = (0 -915928264.8756002) / Math.Sqrt((2100552268231235 / (299)) + (13102130965775.168 / (9))) is greater than 1.9677467375895035 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (9) - 2, .975) and 0.09007743690586592 = (1006600233.9375387 - 915928264.8756002) / 1006600233.9375387 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchF.MDInProd

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 1.52 secs 1.33 secs 0.88 0.00 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchF.MDInProd*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchF.MDInProd.Test

Description of detection logic

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 1.3346294435384614 < 1.4378234977942306.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 196.76574661647675 (T) = (0 -1341012413.5492716) / Math.Sqrt((59360445186986.29 / (299)) + (26360601836324.215 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.11500212572148966 = (1515271903.4975362 - 1341012413.5492716) / 1515271903.4975362 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchI.MDLogicArray

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 354.44 ms 294.00 ms 0.83 0.01 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchI.MDLogicArray*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchI.MDLogicArray.Test

Description of detection logic

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 294.0012206666667 < 336.4827557410714.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 113.24481335470121 (T) = (0 -293915767.34662414) / Math.Sqrt((84702558871754.42 / (299)) + (72768919868.9163 / (44))) is greater than 1.9669451293272056 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (44) - 2, .975) and 0.1705860368994194 = (354365588.7439911 - 293915767.34662414) / 354365588.7439911 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

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

Run Information

Architecture x64
OS ubuntu 18.04
Baseline 338db1a1ba498f308797ce7a54c0ed54e7286d50
Compare ffcb0149763663a55a79784be216afe81f6bf76c
Diff Diff

Improvements in Benchstone.MDBenchF.MDRomber

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Test - Duration of single invocation 181.94 ms 163.02 ms 0.90 0.03 False

graph
Test Report

Repro

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net6.0 --filter 'Benchstone.MDBenchF.MDRomber*'

Payloads

Baseline
Compare

Histogram

Benchstone.MDBenchF.MDRomber.Test

Description of detection logic

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 163.01615503333332 < 169.1859327204762.
IsChangePoint: Marked as a change because one of 7/5/2022 7:01:28 PM, 7/12/2022 6:44:07 AM falls between 7/3/2022 4:26:07 AM and 7/12/2022 6:44:07 AM.
IsImprovementStdDev: Marked as improvement because 26.523036514383964 (T) = (0 -164940727.7717664) / Math.Sqrt((3685055286333.8438 / (299)) + (11474443400035.455 / (45))) is greater than 1.9669246454804268 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (299) + (45) - 2, .975) and 0.07675745411891197 = (178653733.5260657 - 164940727.7717664) / 178653733.5260657 is greater than 0.05.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@tannergooding tannergooding

Labels
arch-x64 branch-refs/heads/main kind-micro os-linux perf-improvement runtime-coreclr
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@EgorBo @tannergooding