[Perf] Linux/x64: 8 Improvements on 7/10/2024 10:56:31 PM

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	4a1a07666b55ed425a5c43896e7869227fca500c
Compare	8ba8249272917366e2382bb4c67a2347d19d2fb6
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
Round - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	21.83 μs	11.79 μs	0.54	0.11	True
CosineSimilarity - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	6.27 μs	1.39 μs	0.22	0.08	False
CosineSimilarity - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	279.20 ns	75.99 ns	0.27	0.13	False
Truncate - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	24.34 μs	11.68 μs	0.48	0.22	True
Round - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	531.06 μs	287.22 μs	0.54	0.13	True
Truncate - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	555.30 μs	285.33 μs	0.51	0.20	True

Test Report

Repro

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

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>*'

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Round(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.CosineSimilarity(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.CosineSimilarity(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Truncate(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Round(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Truncate(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

Docs

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

---

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	4a1a07666b55ed425a5c43896e7869227fca500c
Compare	8ba8249272917366e2382bb4c67a2347d19d2fb6
Diff	Diff
Configs	AOT:true, CompilationMode:wasm, RunKind:micro

Improvements in System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
Truncate - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	14.76 μs	6.03 μs	0.41	0.15	True
Truncate - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	367.97 μs	137.70 μs	0.37	0.18	True

Test Report

Repro

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

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>*'

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Truncate(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Truncate(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

Docs

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

[lewing]

Probably from dotnet/runtime#104651 cc @stephentoub

[stephentoub]

Probably from dotnet/runtime#104651 cc @stephentoub

While I appreciate the credit, this is very unlikely; that PR only added new API and didn't touch any existing implementation.

[lewing]

Probably from dotnet/runtime#104651 cc @stephentoub

While I appreciate the credit, this is very unlikely; that PR only added new API and didn't touch any existing implementation.

range was off in the diff dotnet/runtime@5795e8c...4a1a076 -> dotnet/runtime#103837 for this and a couple of other issue... very odd

[stephentoub]

dotnet/runtime#103837

Yeah, that makes much more sense. Thanks.