Test SIMD 512, 256, 128 registers for fast aggregate calculations.
Unfortunately my hardware doesn't support Vector512.
Anyway, the performance improvement is mindblowing.
Important
net8 is x146 times faster than net48 for calculate the Min and Max at the same time !!
- BenchmarkDotNet=v0.13.5, OS=Windows 10 (10.0.19044.3086/21H2/November2021Update)
- AMD Ryzen 7 1700, 1 CPU, 16 logical and 8 physical cores
- .NET SDK=8.0.100-rc.1.23455.8
- [Host] : .NET 8.0.0 (8.0.23.41904), X64 RyuJIT AVX2
- .NET 7.0 : .NET 7.0.11 (7.0.1123.42427), X64 RyuJIT AVX2
- .NET 8.0 : .NET 8.0.0 (8.0.23.41904), X64 RyuJIT AVX2
- .NET Framework 4.8 : .NET Framework 4.8 (4.8.4644.0), X64 RyuJIT VectorSize=256
| Method | Runtime | Size | Mean | Allocated |
|---|---|---|---|---|
| MinMaxLinq | .NET Framework 4.8 | 10000 | 118,675.226 ns | 65 B |
| MinMaxLinq | .NET 7.0 | 10000 | 2,350.046 ns | - |
| MinMaxLinq | .NET 8.0 | 10000 | 1,228.518 ns | - |
| MinMaxSimd | .NET 7.0 | 10000 | 834.291 ns | - |
| MinMaxSimd | .NET 8.0 | 10000 | 808.150 ns | - |