.NET 7.0 RC1 Microbenchmarks Performance Study Report #76320
Labels
area-Meta
tenet-performance
Performance related issue
tenet-performance-benchmarks
Issue from performance benchmark
tracking
This issue is tracking the completion of other related issues.
Milestone
Comments
ghost
added
the
|
I couldn't figure out the best area label to add to this issue. If you have write-permissions please help me learn by adding exactly one area label. |
EgorBo
added
area-Meta
tenet-performance
|
Tagging subscribers to this area: @dotnet/area-meta Issue DetailsDataThis time we have covered following configurations comparing .NET 7.0 Preview RC1 vs. .NET 7.0 Preview 7:
Most of the benchmarks were run on bare-metal machines, some were executed via WSL. This would not be possible without the help from: @adamsitnik, @AndyAyersMS, @brianrob, @dakersnar, @jkotas, @kunalspathak who contributed their results and time. An addtional thank you to @dakersnar @mrsharm and @jozkee for shadowing the creation of this report and helping with the vetting of the results! The full report generated by the tool is available here. You will have to click "Raw" to see the entire file. The report is sorted from most regressed to most improved, so scroll to the bottom in the full report to see improvements. There are plenty of them! Again, the full historical data turned out to be extremely useful. For details about methodology please read #41871. Preview 7 report can be found here. Notable Improvements
Regressions
By Design
Already Solved
Noise, Flaky or MultimodalA lot of benchmarks showed up in the report generated by the tool, but were not actual regressions, they mostly are the same as in the previous report Some regressions looked real but turned out to be stale PGO related and fixed with PGO updates, e.g. #72866 StatisticsTotal: 52498 Statistics per Architecture
Statistics per Operating System
Statistics per Namespace
Big thanks to everyone involved!
|
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Data
This time we have covered following configurations comparing .NET 7.0 Preview RC1 vs. .NET 7.0 Preview 7:
Most of the benchmarks were run on bare-metal machines, some were executed via WSL.
This would not be possible without the help from: @adamsitnik, @AndyAyersMS, @brianrob, @dakersnar, @jkotas, @kunalspathak who contributed their results and time. An addtional thank you to @dakersnar @mrsharm and @jozkee for shadowing the creation of this report and helping with the vetting of the results!
The full report generated by the tool is available here. You will have to click "Raw" to see the entire file. The report is sorted from most regressed to most improved, so scroll to the bottom in the full report to see improvements. There are plenty of them!
Again, the full historical data turned out to be extremely useful. For details about methodology please read #41871. Preview 7 report can be found here.
Notable Improvements
System.Memory.Span<Char>.LastIndexOfAnyValues(Size: 512),System.Collections.ContainsFalse<Int32>.*,System.Tests.Perf_Array.IndexOfShort,System.Tests.Perf_String.Replace_Charand other benchmarksSystem.Perf_Convert.ToBase64CharArray(binaryDataSize: 1024, formattingOptions: None)System.Tests.Perf_Environment.GetLogicalDrivesSystem.Text.Json.Serialization.Tests.ReadJson<ArrayList>.DeserializeFromUtf8BytesSystem.Text.RegularExpressions.Tests.Perf_Regex_Industry_RustLang_Sherlock.Count(Pattern: "/w+/s+Holmes/s+/w+", Options: NonBacktracking)System.Reflection.Attributes.GetCustomAttributes*Regressions
A lot of regressions such as
System.Tests.Perf_Enum.GetNamefrom "More precise writebarriers for Regions"System.Text.RegularExpressions.Tests.Perf_Regex_Industry_RustLang_Sherlock.Count(),System.Globalization.Tests.StringSearch.IndexOf_Word_NotFound(Options: (en-US, Ordinal, False))System.Text.Json.Tests.Utf8JsonReaderCommentsTestson arm64*.CastingPerf.*benchmarksSystem.IO.Compression.GzipBy Design
Various improvements and regressions for
System.IO.Compression.Brotli.*(level:Optimal, *)CompressionLevel.Optimalwas changed in Fix CompressionLevel.Optimal for Brotli #72266Various regressions in XmlSerializer, e.g.
MicroBenchmarks.Serializers.Xml_ToStream<MyEventsListerViewModel>.XmlSerializer_System.Collections.Concurrent.IsEmpty<String>.Dictionary(Size: 512)System.IO.Tests.StreamReaderReadToEndTests.ReadToEndAsyncand other GC-intensive benchmarksAlready Solved
String.ReplaceSystem.Numerics.Tests.Perf_BigInteger.ParseNoise, Flaky or Multimodal
A lot of benchmarks showed up in the report generated by the tool, but were not actual regressions, they mostly are the same as in the previous report
Some regressions looked real but turned out to be stale PGO related and fixed with PGO updates, e.g. #72866
Statistics
Total: 52498
Same: 72.28 %
Slower: 10.49 %
Faster: 8.19 %
Noise: 8.97 %
Unknown: 0.07 %
Statistics per Architecture
Statistics per Operating System
Statistics per Namespace
Big thanks to everyone involved!
The text was updated successfully, but these errors were encountered: