Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Improve Ascii.FromUtf16 and FromHexString #102735

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
Merged
merged 2 commits into from
May 29, 2024
Merged

Improve Ascii.FromUtf16 and FromHexString #102735

merged 2 commits into from
May 29, 2024

Conversation

MihaZupan
Copy link
Member

@MihaZupan MihaZupan commented May 27, 2024 •
edited
Loading

We're already using things like Sse2.PackUnsignedSaturate instead of Vector128.Narrow when we don't care about the edge case values (we've already validated the input).
This PR does something similar for Vector256 and Vector512 paths.

public class AsciiBenchmarks
{
    [Params(1000)]
    public int Length;

    private byte[] _bytes;
    private char[] _chars;

    [GlobalSetup]
    public void Setup()
    {
        _chars = new string('a', Length).ToCharArray();
        _bytes = Encoding.ASCII.GetBytes(_chars);
    }

    [Benchmark]
    public OperationStatus FromUtf16() => Ascii.FromUtf16(_chars, _bytes, out _);

    [Benchmark]
    public OperationStatus FromHexString() => Convert.FromHexString(_chars, _bytes, out _, out _);
}
Method Toolchain Length Mean Error Ratio
FromUtf16 main 1000 45.37 ns 0.057 ns 1.00
FromUtf16 pr 1000 32.70 ns 0.163 ns 0.72
FromHexString main 1000 144.48 ns 0.097 ns 1.00
FromHexString pr 1000 102.58 ns 0.079 ns 0.71

(this is with Avx2)

Diffs: https://gist.github.com/MihuBot/f60c7254060f436ae9c12ded701c85f8

@MihaZupan MihaZupan added this to the 9.0.0 milestone May 27, 2024
@dotnet-policy-service Dotnet Policy Service
Copy link
Contributor

Tagging subscribers to this area: @dotnet/area-system-text-encoding
See info in area-owners.md if you want to be subscribed.

@EgorBo
Copy link
Member

EgorBo commented May 28, 2024

@EgorBot -intel -amd --disasm

using BenchmarkDotNet.Attributes;
using System.Buffers;
using System.Text;
using System;
using BenchmarkDotNet.Running;

BenchmarkRunner.Run<AsciiBenchmarks>(args: args);

public class AsciiBenchmarks
{
    [Params(1000)]
    public int Length;

    private byte[] _bytes;
    private char[] _chars;

    [GlobalSetup]
    public void Setup()
    {
        _chars = new string('a', Length).ToCharArray();
        _bytes = Encoding.ASCII.GetBytes(_chars);
    }

    [Benchmark]
    public OperationStatus FromUtf16() => Ascii.FromUtf16(_chars, _bytes, out _);
}

@EgorBot
Copy link

EgorBot commented May 28, 2024 

BenchmarkDotNet v0.13.12, Ubuntu 22.04.4 LTS (Jammy Jellyfish)
Intel Xeon Platinum 8370C CPU 2.80GHz, 1 CPU, 16 logical and 8 physical cores
  Job-BWGLXM : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX-512F+CD+BW+DQ+VL+VBMI
  Job-ULJSQF : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX-512F+CD+BW+DQ+VL+VBMI
Method Toolchain Length Mean Error Ratio Code Size
FromUtf16 Main 1000 34.96 ns 0.007 ns 1.00 1,104 B
FromUtf16 PR 1000 27.99 ns 0.006 ns 0.80 1,033 B
More data

See BDN_Artifacts.zip for details.

@EgorBot
Copy link

EgorBot commented May 28, 2024 

BenchmarkDotNet v0.13.12, Ubuntu 22.04.4 LTS (Jammy Jellyfish)
AMD EPYC 7763, 1 CPU, 16 logical and 8 physical cores
  Job-NUESEZ : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX2
  Job-WIGDOZ : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX2
Method Toolchain Length Mean Error Ratio Code Size
FromUtf16 Main 1000 33.02 ns 0.004 ns 1.00 796 B
FromUtf16 PR 1000 23.83 ns 0.006 ns 0.72 760 B
More data

See BDN_Artifacts.zip for details.

@ShreyasJejurkar
Copy link
Contributor

Whooooo nice bot @EgorBo. Pretty interesting...😅🧐

tannergooding
tannergooding reviewed May 28, 2024
View reviewed changes
@MihaZupan
Copy link
Member Author

@EgorBot -arm64 -amd --disasm

using BenchmarkDotNet.Attributes;
using System.Buffers;
using System.Text;
using System;
using BenchmarkDotNet.Running;

BenchmarkRunner.Run<AsciiBenchmarks>(args: args);

public class AsciiBenchmarks
{
    [Params(1000)]
    public int Length;

    private byte[] _bytes;
    private char[] _chars;

    [GlobalSetup]
    public void Setup()
    {
        _chars = new string('a', Length).ToCharArray();
        _bytes = Encoding.ASCII.GetBytes(_chars);
    }

    [Benchmark]
    public OperationStatus FromHexString() => Convert.FromHexString(_chars, _bytes, out _, out _);
}

@EgorBot
Copy link

EgorBot commented May 28, 2024 •
edited by EgorBo
Loading 

BenchmarkDotNet v0.13.12, Ubuntu 22.04.4 LTS (Jammy Jellyfish)
Intel Xeon Platinum 8370C CPU 2.80GHz, 1 CPU, 16 logical and 8 physical cores
  Job-KKQGMF : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX-512F+CD+BW+DQ+VL+VBMI
  Job-GAATJR : .NET 9.0.0 (42.42.42.42424), X64 RyuJIT AVX-512F+CD+BW+DQ+VL+VBMI
Method Toolchain Length Mean Error Ratio
FromHexString Main 1000 131.7 ns 0.11 ns 1.00
FromHexString PR 1000 115.3 ns 0.02 ns 0.88
More data

See BDN_Artifacts.zip for details.

@dotnet dotnet deleted a comment from EgorBot May 28, 2024
@tannergooding tannergooding merged commit 5c74f63 into dotnet:main May 29, 2024
Ruihan-Yin pushed a commit to Ruihan-Yin/runtime that referenced this pull request May 30, 2024
@MihaZupan @Ruihan-Yin
Improve Ascii.FromUtf16 and FromHexString (dotnet#102735)
a60f57f 
* Use cheaper Vector.Narrow altenatives in Ascii

* Use existing Ascii helper in HexConverter
@github-actions github-actions bot locked and limited conversation to collaborators Jun 29, 2024
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Reviewers

@tannergooding tannergooding tannergooding approved these changes

Assignees

@MihaZupan MihaZupan

Labels
area-System.Text.Encoding
Projects
None yet
Milestone
9.0.0
Development

Successfully merging this pull request may close these issues.
5 participants
@MihaZupan @EgorBo @EgorBot @ShreyasJejurkar @tannergooding