youmuliu/stgsharp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Daily update on 20260111-005714 by Nitload Youmu. Additional Message : Use optimized tile operation to accelerate matrix mul, still has problem

Browse Source 
Signed-off-by: Nitload <YoumuLiu_NSI@whv0t.onmicrosoft.com>
This commit is contained in:
Nitload
2026-01-11 00:57:14 +08:00
parent 6827259891
commit 4be5c1c7c9
9 changed files with 141 additions and 219 deletions
Download Patch File Download Diff File Expand all files Collapse all files
StgSharp.Mathematics/Numeric/MatrixCompute.Float.cs
+28 -44
View File
@@ -1,9 +1,9 @@
//-----------------------------------------------------------------------
//-----------------------------------------------------------------------
// -----------------------------------------------------------------------
// file="MatrixCompute.Float"
// Project: StgSharp
// AuthorGroup: Nitload
// Copyright (c) Nitload. All rights reserved.
// AuthorGroup: Nitload Space
// Copyright (c) Nitload Space. All rights reserved.
//
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to deal
@@ -42,11 +42,7 @@ namespace StgSharp.Mathematics.Numeric
public static partial class MatrixCompute
{
public static unsafe void Add(
Matrix<float> left,
Matrix<float> right,
Matrix<float> ans
)
public static unsafe void Add(Matrix<float> left, Matrix<float> right, Matrix<float> ans)
{
int f32 = (int)MatrixElementType.F32;
if (left.ColumnLength != right.ColumnLength ||
@@ -92,10 +88,7 @@ namespace StgSharp.Mathematics.Numeric
return;
}
public static unsafe void Fill(
Matrix<float> ans,
float value
)
public static unsafe void Fill(Matrix<float> ans, float value)
{
int f32 = (int)MatrixElementType.F32;
long count = (long)ans.KernelColumnLength * ans.KernelRowLength;
@@ -126,11 +119,7 @@ namespace StgSharp.Mathematics.Numeric
taskGroup.Dispose();
}
public static unsafe void Mul(
Matrix<float> left,
Matrix<float> right,
Matrix<float> ans
)
public static unsafe void Mul(Matrix<float> left, Matrix<float> right, Matrix<float> ans)
{
int f32 = (int)MatrixElementType.F32;
@@ -155,28 +144,32 @@ namespace StgSharp.Mathematics.Numeric
if (count > 256) {
pool = MatrixParallel.LeadParallel(1);
}
M128* enumeration = stackalloc M128[4];
for (int i = 0; i < ans.KernelRowLength; i++)
{
for (int j = 0; j < ans.KernelColumnLength; j++)
{
MatrixKernel<float>* ansPanel = GetKernelAddressUnsafe(ans.Buffer,
ans.KernelColumnLength,
i, j);
enumeration->Member<int>(0) = i;
enumeration->Member<int>(1) = j;
enumeration->Member<int>(2) = ans.KernelRowLength;
enumeration->Member<int>(3) = left.KernelColumnLength;
// NativeIntrinsic.Context.mat[f32].clear_panel((MatrixPanel*)ansPanel);
for (int k = 0; k < left.KernelColumnLength; k++)
{
NativeIntrinsic.Context
.mat[f32].kernel_tile_fma((MatrixKernel*)left.Buffer,
(MatrixKernel*)right.Buffer,
(MatrixKernel*)ans.Buffer, enumeration);
/*
for (int k = 0; k < left.KernelColumnLength; k++) {
NativeIntrinsic.Context
.mat[f32].kernel_fma((MatrixPanel*)GetKernelAddressUnsafe(left.Buffer,
left.KernelColumnLength,
i,
k),
(MatrixPanel*)GetKernelAddressUnsafe(right.Buffer,
right.KernelColumnLength,
k,
j),
(MatrixPanel*)ansPanel);
.mat[f32].kernel_fma(GetKernelAddressUnsafe(left.Buffer,
left.KernelColumnLength, i,
k),
GetKernelAddressUnsafe(right.Buffer,
right.KernelColumnLength, k, j),
GetKernelAddressUnsafe(ans.Buffer,
ans.KernelColumnLength, i, j));
}
/**/
}
}
@@ -203,18 +196,13 @@ namespace StgSharp.Mathematics.Numeric
return;
}
public static unsafe void Mul(
Matrix<float> left,
float right,
Matrix<float> ans
)
public static unsafe void Mul(Matrix<float> left, float right, Matrix<float> ans)
{
int f32 = (int)MatrixElementType.F32;
if (left.ColumnLength != ans.ColumnLength || left.RowLength != ans.RowLength) {
throw new IndexOutOfRangeException("Matrix dimensions must agree.");
}
if (left.Layout != MatrixLayout.DenseRectangle ||
ans.Layout != MatrixLayout.DenseRectangle) {
if (left.Layout != MatrixLayout.DenseRectangle || ans.Layout != MatrixLayout.DenseRectangle) {
throw new NotSupportedException("Only dense rectangle matrices are supported.");
}
long count = (long)left.KernelColumnLength * left.KernelRowLength;
@@ -255,11 +243,7 @@ namespace StgSharp.Mathematics.Numeric
return;
}
public static unsafe void Sub(
Matrix<float> left,
Matrix<float> right,
Matrix<float> ans
)
public static unsafe void Sub(Matrix<float> left, Matrix<float> right, Matrix<float> ans)
{
int f32 = (int)MatrixElementType.F32;
if (left.ColumnLength != right.ColumnLength ||
StgSharp.Mathematics/Numeric/MatrixCompute.cs
+27 -28
View File
@@ -1,9 +1,9 @@
//-----------------------------------------------------------------------
//-----------------------------------------------------------------------
// -----------------------------------------------------------------------
// file="MatrixCompute"
// Project: StgSharp
// AuthorGroup: Nitload
// Copyright (c) Nitload. All rights reserved.
// AuthorGroup: Nitload Space
// Copyright (c) Nitload Space. All rights reserved.
//
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to deal
@@ -40,9 +40,7 @@ namespace StgSharp.Mathematics.Numeric
public partial class MatrixCompute
{
internal static unsafe void ComputeFmaPanelF32(
MatrixParallelTaskPackage* p
)
internal static unsafe void ComputeFmaPanelF32(MatrixParallelTaskPackage* p)
{
int f32 = (int)MatrixElementType.F32;
MatrixKernel* leftBuffer = (MatrixKernel*)p->Left;
@@ -64,19 +62,14 @@ namespace StgSharp.Mathematics.Numeric
for (int k = 0; k < commonK; k++) {
NativeIntrinsic.Context
.mat[f32].kernel_fma(
(MatrixPanel*)GetKernelAddressUnsafe((MatrixKernel<float>*)leftBuffer,
ansHeight, k, j),
(MatrixPanel*)GetKernelAddressUnsafe((MatrixKernel<float>*)rightBuffer,
commonK, j, k),
(MatrixPanel*)GetKernelAddressUnsafe((MatrixKernel<float>*)ansBuffer,
ansHeight, i, j));
GetKernelAddressUnsafe((MatrixKernel<float>*)leftBuffer, ansHeight, k, j),
GetKernelAddressUnsafe((MatrixKernel<float>*)rightBuffer, commonK, j, k),
GetKernelAddressUnsafe((MatrixKernel<float>*)ansBuffer, ansHeight, i, j));
}
}
}
internal static unsafe void ComputeFmaPanelF64(
MatrixParallelTaskPackage* p
)
internal static unsafe void ComputeFmaPanelF64(MatrixParallelTaskPackage* p)
{
int f64 = (int)MatrixElementType.F64;
@@ -90,33 +83,39 @@ namespace StgSharp.Mathematics.Numeric
int ansHeight = p->AnsSecOffset;
int commonK = p->LeftPrimOffset;
M128* enumeration = stackalloc M128[1];
// in this occasion size must be 1
for (long c = offsetRef; c < offsetRef + lengthRef; c++)
{
int i = (int)(c / ansWidth);
int j = (int)(c % ansWidth);
enumeration->Member<int>(0) = i;
enumeration->Member<int>(1) = j;
enumeration->Member<int>(2) = commonK;
enumeration->Member<int>(3) = ansHeight;
NativeIntrinsic.Context.mat[f64].kernel_tile_fma(leftBuffer, rightBuffer, ansBuffer, enumeration);
/*
for (int k = 0; k < commonK; k += 1) {
NativeIntrinsic.Context
.mat[f64].kernel_fma(
(MatrixPanel*)GetKernelAddressUnsafe((MatrixKernel<double>*)leftBuffer,
ansHeight, k, j),
(MatrixPanel*)GetKernelAddressUnsafe((MatrixKernel<double>*)rightBuffer,
commonK, j, k),
(MatrixPanel*)GetKernelAddressUnsafe((MatrixKernel<double>*)ansBuffer,
ansHeight, i, j));
GetKernelAddressUnsafe((MatrixKernel<double>*)leftBuffer, ansHeight, k, j),
GetKernelAddressUnsafe((MatrixKernel<double>*)rightBuffer, commonK, j, k),
GetKernelAddressUnsafe((MatrixKernel<double>*)ansBuffer, ansHeight, i, j));
}
/**/
}
}
[MethodImpl(MethodImplOptions.AggressiveInlining)]
internal static unsafe MatrixKernel<T>* GetKernelAddressUnsafe<T>(
MatrixKernel<T>* buffer,
int colLength,
int x,
int y
) where T : unmanaged, INumber<T>
internal static unsafe MatrixKernel* GetKernelAddressUnsafe<T>(
MatrixKernel<T>* buffer,
int colLength,
int x,
int y)
where T: unmanaged, INumber<T>
{
return buffer + (colLength * x + y);
return (MatrixKernel*)(buffer + (colLength * x + y));
}
}
StgSharp.Mathematics/Numeric/MatrixCompute.double.cs
+35 -38
View File
@@ -1,9 +1,9 @@
//-----------------------------------------------------------------------
//-----------------------------------------------------------------------
// -----------------------------------------------------------------------
// file="MatrixCompute.double"
// Project: StgSharp
// AuthorGroup: Nitload
// Copyright (c) Nitload. All rights reserved.
// AuthorGroup: Nitload Space
// Copyright (c) Nitload Space. All rights reserved.
//
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to deal
@@ -38,11 +38,7 @@ namespace StgSharp.Mathematics.Numeric
public partial class MatrixCompute
{
public static unsafe void Add(
Matrix<double> left,
Matrix<double> right,
Matrix<double> ans
)
public static unsafe void Add(Matrix<double> left, Matrix<double> right, Matrix<double> ans)
{
int f64 = (int)MatrixElementType.F64;
if (left.ColumnLength != right.ColumnLength ||
@@ -88,10 +84,7 @@ namespace StgSharp.Mathematics.Numeric
return;
}
public static unsafe void Fill(
Matrix<double> ans,
double value
)
public static unsafe void Fill(Matrix<double> ans, double value)
{
int f64 = (int)MatrixElementType.F64;
long count = (long)ans.KernelColumnLength * ans.KernelRowLength;
@@ -122,11 +115,7 @@ namespace StgSharp.Mathematics.Numeric
taskGroup.Dispose();
}
public static unsafe void Mul(
Matrix<double> left,
Matrix<double> right,
Matrix<double> ans
)
public static unsafe void Mul(Matrix<double> left, Matrix<double> right, Matrix<double> ans)
{
int f64 = (int)MatrixElementType.F64;
@@ -139,25 +128,42 @@ namespace StgSharp.Mathematics.Numeric
throw new IndexOutOfRangeException("Matrix dimensions must agree.");
}
long count = (long)left.KernelColumnLength * left.KernelRowLength;
int size = MatrixPanel.KernelSideCount<double>();
// int size = MatrixPanel.KernelSideCount<double>();
if (count <= 1024)
{
MatrixParallelThread[] pool = null;
if (count > 256) {
pool = MatrixParallel.LeadParallel(1);
}
for (int i = 0; i < ans.KernelRowLength; i += size)
M128* enumeration = stackalloc M128[1];
for (int i = 0; i < ans.KernelRowLength; i++)
{
for (int j = 0; j < ans.KernelColumnLength; j += size)
for (int j = 0; j < ans.KernelColumnLength; j += 1)
{
enumeration->Member<int>(0) = i;
enumeration->Member<int>(1) = j;
enumeration->Member<int>(2) = ans.KernelRowLength;
enumeration->Member<int>(3) = left.KernelColumnLength;
NativeIntrinsic.Context
.mat[f64].kernel_tile_fma((MatrixKernel*)left.Buffer,
(MatrixKernel*)right.Buffer,
(MatrixKernel*)ans.Buffer, enumeration);
/*
// Console.WriteLine($"{i}{j}");
for (int k = 0; k < left.KernelColumnLength; k += size)
for (int k = 0; k < left.KernelColumnLength; k++)
{
NativeIntrinsic.Context
.mat[f64].kernel_fma((MatrixPanel*)left.GetKernelAddressUnsafe(i, k),
(MatrixPanel*)right.GetKernelAddressUnsafe(k, j),
(MatrixPanel*)ans.GetKernelAddressUnsafe(i, j));
.mat[f64].kernel_fma(GetKernelAddressUnsafe(left.Buffer,
left.KernelColumnLength, i,
k),
GetKernelAddressUnsafe(right.Buffer,
right.KernelColumnLength, k, j),
GetKernelAddressUnsafe(ans.Buffer,
ans.KernelColumnLength, i, j));
}
/**/
}
}
if (count > 256) {
@@ -172,8 +178,8 @@ namespace StgSharp.Mathematics.Numeric
package->AnsPrimOffset = ans.KernelRowLength;
package->AnsSecOffset = ans.KernelColumnLength;
package->LeftPrimOffset = left.KernelColumnLength;
package->PrimCount = ans.KernelColumnLength / size;
package->SecCount = ans.KernelRowLength / size;
package->PrimCount = ans.KernelColumnLength;
package->SecCount = ans.KernelRowLength;
package->ElementSize = (int)MatrixElementType.F64;
package->ComputeMode = new MatrixOpMode(MatrixIndexStyle.BUFFER_INDEX, MatrixOperationSpecial.SEQ_FMA);
MatrixParallelHandle taskGroup = MatrixParallel.ScheduleTask(package);
@@ -183,18 +189,13 @@ namespace StgSharp.Mathematics.Numeric
return;
}
public static unsafe void Mul(
Matrix<double> left,
double right,
Matrix<double> ans
)
public static unsafe void Mul(Matrix<double> left, double right, Matrix<double> ans)
{
int f64 = (int)MatrixElementType.F64;
if (left.ColumnLength != ans.ColumnLength || left.RowLength != ans.RowLength) {
throw new IndexOutOfRangeException("Matrix dimensions must agree.");
}
if (left.Layout != MatrixLayout.DenseRectangle ||
ans.Layout != MatrixLayout.DenseRectangle) {
if (left.Layout != MatrixLayout.DenseRectangle || ans.Layout != MatrixLayout.DenseRectangle) {
throw new NotSupportedException("Only dense rectangle matrices are supported.");
}
long count = (long)left.KernelColumnLength * left.KernelRowLength;
@@ -235,11 +236,7 @@ namespace StgSharp.Mathematics.Numeric
return;
}
public static unsafe void Sub(
Matrix<double> left,
Matrix<double> right,
Matrix<double> ans
)
public static unsafe void Sub(Matrix<double> left, Matrix<double> right, Matrix<double> ans)
{
int f64 = (int)MatrixElementType.F64;
if (left.ColumnLength != right.ColumnLength ||