[Perf] Windows/arm64: 11 Improvements on 7/10/2024 2:35:18 PM

[performanceautofiler]

Run Information

Name	Value
Architecture	arm64
OS	Windows 10.0.22621
Queue	SurfaceWindows
Baseline	9882188fbe16adee3538b65a3926465dfd6052c1
Compare	0d426df37c6ac2c1d22f952a2e359ffd4281d47b
Diff	Diff
Configs	CompilationMode:tiered, 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
Exp - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	695.14 ns	500.93 ns	0.72	0.01	True
Sigmoid - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	18.39 μs	13.67 μs	0.74	0.01	True
Log - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	925.37 ns	620.27 ns	0.67	0.01	True
Round - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	10.80 μs	672.48 ns	0.06	0.07	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
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>*'

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

ETL Files

Histogram

JIT Disasms

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

ETL Files

Histogram

JIT Disasms

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

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Round(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	arm64
OS	Windows 10.0.22621
Queue	SurfaceWindows
Baseline	4addcaa7e36104390835355c5a98eb36ec0334d0
Compare	635f9af977bef698e8092d51386d3ab546395e0a
Diff	Diff
Configs	CompilationMode:tiered, 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
Round - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	68.29 ns	30.39 ns	0.44	0.02	True
FusedMultiplyAdd_Vectors - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	1.34 μs	680.35 ns	0.51	0.03	True
Log - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	354.49 ns	248.32 ns	0.70	0.02	False
Sigmoid - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	405.46 ns	335.24 ns	0.83	0.01	True
Truncate - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	1.38 μs	532.53 ns	0.39	0.15	True
Round - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	1.40 μs	531.64 ns	0.38	0.01	True
FusedMultiplyAdd_ScalarMultiplier - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	1.51 μs	497.65 ns	0.33	0.02	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
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>*'

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

ETL Files

Histogram

JIT Disasms

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

ETL Files

Histogram

JIT Disasms

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

ETL Files

Histogram

JIT Disasms

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

ETL Files

Histogram

JIT Disasms

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

ETL Files

Histogram

JIT Disasms

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

ETL Files

Histogram

JIT Disasms

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

ETL Files

Histogram

JIT Disasms

Docs

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

[LoopedBard3]

Improvement likely caused by: dotnet/runtime#103837