mlemanczyk/even-perfect-numbers-scanner
1
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Optimizations

Browse Source
This commit is contained in:
Marek Lemańczyk committed 2025-09-24 00:01:08 +02:00
1 parent 7079d108e7
commit 773848c577
1 file changed
+46 -28
Download Patch File Download Diff File Expand all files Collapse all files
PerfectNumbers.Core/Gpu/MersenneNumberDivisorByDivisorGpuTester.cs
+46 -28
View File
@@ -318,27 +318,45 @@ public sealed class MersenneNumberDivisorByDivisorGpuTester
int offset = 0;
int consoleStatus = 0;
bool useCycles = _owner._useDivisorCycles;
MersenneNumberDivisorByDivisorGpuTester owner = _owner;
int gpuBatchSize = owner._gpuBatchSize;
bool useCycles = owner._useDivisorCycles;
byte[] hitsHost = _hitsHost;
int[] indexHost = _indexHost;
ulong[] primesHost = _primesHost;
Accelerator accelerator = _accelerator;
MemoryBuffer1D<byte, Stride1D.Dense> hitsBuffer = _hitsBuffer;
MemoryBuffer1D<ulong, Stride1D.Dense> primesBuffer = _primesBuffer;
ArrayView1D<ulong, Stride1D.Dense> primesBufferView = primesBuffer.View;
ArrayView1D<byte, Stride1D.Dense> hitsBufferView = hitsBuffer.View;
Action<Index1D, MontgomeryDivisorData, ArrayView<ulong>, ArrayView<byte>> kernel = _kernel;
ArrayView1D<ulong, Stride1D.Dense> primesView;
ArrayView1D<byte, Stride1D.Dense> hitsView;
MersenneDivisorCycles? divisorCycles = useCycles ? MersenneDivisorCycles.Shared : null;
ulong cycle = useCycles ? divisorCycles!.GetCycle(divisor) : 0UL;
ulong cycle = useCycles ? divisorCycles!.GetCycle(divisor) : 0UL, primeValue;
MontgomeryDivisorData divisorData = CreateMontgomeryDivisorData(divisor);
Span<ulong> primeSpan = _primesHost.AsSpan(0, _owner._gpuBatchSize);
Span<int> indexSpan = _indexHost.AsSpan(0, _owner._gpuBatchSize);
Span<byte> hitSpan = _hitsHost.AsSpan(0, _owner._gpuBatchSize);
int currentSpanSize = _owner._gpuBatchSize;
Span<ulong> primeSpan = primesHost.AsSpan(0, gpuBatchSize);
ref ulong primesSpanRef = ref MemoryMarshal.GetReference(primeSpan);
Span<int> indexSpan = indexHost.AsSpan(0, gpuBatchSize);
Span<byte> hitsSlice, hitSpan = hitsHost.AsSpan(0, gpuBatchSize);
ref byte hitSpanRef = ref MemoryMarshal.GetReference(hitSpan);
int batchSize, currentSpanSize = gpuBatchSize, gpuCount, i, primesLength = primes.Length;
while (offset < primes.Length)
while (offset < primesLength)
{
int batchSize = Math.Min(_owner._gpuBatchSize, primes.Length - offset);
batchSize = Math.Min(gpuBatchSize, primesLength - offset);
if (batchSize != currentSpanSize)
{
primeSpan = _primesHost.AsSpan(0, batchSize);
indexSpan = _indexHost.AsSpan(0, batchSize);
hitSpan = _hitsHost.AsSpan(0, batchSize);
primeSpan = primesHost.AsSpan(0, batchSize);
primesSpanRef = ref MemoryMarshal.GetReference(primeSpan);
indexSpan = indexHost.AsSpan(0, batchSize);
hitSpan = hitsHost.AsSpan(0, batchSize);
hitSpanRef = ref MemoryMarshal.GetReference(hitSpan);
currentSpanSize = batchSize;
}
primes.Slice(offset, batchSize).CopyTo(primeSpan);
// We benefit from the constructor, because CopyTo will be called as static function
new Span<ulong>(primesHost, offset, batchSize).CopyTo(primeSpan);
if (++consoleStatus == PerfectNumberConstants.ConsoleInterval)
{
@@ -346,14 +364,14 @@ public sealed class MersenneNumberDivisorByDivisorGpuTester
consoleStatus = 0;
}
Span<byte> hitsSlice = hits.Slice(offset, batchSize);
int gpuCount = 0;
hitsSlice = hits.Slice(offset, batchSize);
gpuCount = 0;
if (useCycles)
{
for (int i = 0; i < batchSize; i++)
for (i = 0; i < batchSize; i++)
{
ulong primeValue = primeSpan[i];
primeValue = primeSpan[i];
if (cycle == 0UL || primeValue % cycle != 0UL)
{
hitsSlice[i] = 0;
@@ -367,15 +385,15 @@ public sealed class MersenneNumberDivisorByDivisorGpuTester
if (gpuCount > 0)
{
var primesView = _primesBuffer.View.SubView(0, gpuCount);
var hitsView = _hitsBuffer.View.SubView(0, gpuCount);
primesView = primesBufferView.SubView(0, gpuCount);
hitsView = hitsBufferView.SubView(0, gpuCount);
primesView.CopyFromCPU(ref MemoryMarshal.GetReference(primeSpan), gpuCount);
_kernel(gpuCount, divisorData, primesView, hitsView);
_accelerator.Synchronize();
primesView.CopyFromCPU(ref primesSpanRef, gpuCount);
kernel(gpuCount, divisorData, primesView, hitsView);
accelerator.Synchronize();
hitsView.CopyToCPU(ref MemoryMarshal.GetReference(hitSpan), gpuCount);
for (int i = 0; i < gpuCount; i++)
hitsView.CopyToCPU(ref hitSpanRef, gpuCount);
for (i = 0; i < gpuCount; i++)
{
hitsSlice[indexSpan[i]] = hitSpan[i];
}
@@ -383,12 +401,12 @@ public sealed class MersenneNumberDivisorByDivisorGpuTester
}
else
{
var primesView = _primesBuffer.View.SubView(0, batchSize);
var hitsView = _hitsBuffer.View.SubView(0, batchSize);
primesView = primesBufferView.SubView(0, batchSize);
hitsView = hitsBufferView.SubView(0, batchSize);
primesView.CopyFromCPU(ref MemoryMarshal.GetReference(primeSpan), batchSize);
_kernel(batchSize, divisorData, primesView, hitsView);
_accelerator.Synchronize();
primesView.CopyFromCPU(ref primesSpanRef, batchSize);
kernel(batchSize, divisorData, primesView, hitsView);
accelerator.Synchronize();
hitsView.CopyToCPU(ref MemoryMarshal.GetReference(hitSpan), batchSize);
hitSpan.CopyTo(hitsSlice);