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

Replace small-modulo uses with optimized helpers

Browse Source
This commit is contained in:
Marek Lemańczyk committed 2025-09-22 19:54:50 +02:00
1 parent ca866ffc5b
commit 3598f88aaf
9 files changed
+46 -38
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

PerfectNumbers.Core/Gpu/GpuKernelPool.cs
+22 -14
View File
@@ -200,8 +200,10 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
ArrayView1D<uint, Stride1D.Dense> smallCycles)
{
ulong idx = (ulong)index.X;
// residue automaton
ulong step10 = (exponent.Mod10() << 1).Mod10();
ulong idxMod3 = idx.Mod3();
ulong idxMod5 = idx.Mod5();
// residue automaton
ulong step10 = (exponent.Mod10() << 1).Mod10();
ulong r10 = q0m10 + (step10 * idx).Mod10();
r10 -= (r10 >= 10UL) ? 10UL : 0UL;
bool shouldCheck = r10 != 5UL; // skip q ≡ 5 (mod 10); other residues handled by r8/r3/r5
@@ -215,16 +217,16 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
}
else
{
ulong step3 = ((exponent % 3UL) << 1) % 3UL;
ulong r3 = q0m3 + (step3 * (idx % 3UL)); r3 -= (r3 >= 3UL) ? 3UL : 0UL;
ulong step3 = (exponent.Mod3() << 1).Mod3();
ulong r3 = q0m3 + (step3 * idxMod3); r3 -= (r3 >= 3UL) ? 3UL : 0UL;
if (r3 == 0UL)
{
shouldCheck = false;
}
else
{
ulong step5Loc = ((exponent % 5UL) << 1) % 5UL;
ulong r5 = q0m5 + (step5Loc * (idx % 5UL)); if (r5 >= 5UL) r5 -= 5UL;
ulong step5Loc = (exponent.Mod5() << 1).Mod5();
ulong r5 = q0m5 + (step5Loc * idxMod5); if (r5 >= 5UL) r5 -= 5UL;
if (r5 == 0UL)
{
shouldCheck = false;
@@ -358,6 +360,8 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
ArrayView1D<ulong, Stride1D.Dense> smallPrimesPow2LastSeven)
{
ulong idx = (ulong)index.X;
ulong idxMod3 = idx.Mod3();
ulong idxMod5 = idx.Mod5();
ulong r10 = ra.Q0M10 + (ra.Step10 * idx).Mod10(); r10 -= (r10 >= 10UL) ? 10UL : 0UL;
bool shouldCheck = r10 != 5UL;
if (shouldCheck)
@@ -369,7 +373,7 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
}
else
{
ulong r3 = ra.Q0M3 + (ra.Step3 * (idx % 3UL));
ulong r3 = ra.Q0M3 + (ra.Step3 * idxMod3);
if (r3 >= 6UL) r3 -= 6UL;
if (r3 >= 3UL) r3 -= 3UL;
if (r3 == 0UL)
@@ -378,7 +382,7 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
}
else
{
ulong r5 = ra.Q0M5 + (ra.Step5 * (idx % 5UL));
ulong r5 = ra.Q0M5 + (ra.Step5 * idxMod5);
if (r5 >= 15UL) r5 -= 15UL;
if (r5 >= 10UL) r5 -= 10UL;
if (r5 >= 5UL) r5 -= 5UL;
@@ -489,7 +493,9 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
ResidueAutomatonArgs ra, ArrayView<int> found, ArrayView1D<uint, Stride1D.Dense> smallCycles)
{
ulong idx = (ulong)index.X;
ulong r10 = ra.Q0M10 + (ra.Step10 * idx).Mod10(); r10 -= (r10 >= 10UL) ? 10UL : 0UL;
ulong idxMod3 = idx.Mod3();
ulong idxMod5 = idx.Mod5();
ulong r10 = ra.Q0M10 + (ra.Step10 * idx).Mod10(); r10 -= (r10 >= 10UL) ? 10UL : 0UL;
bool shouldCheck = r10 != 5UL;
if (shouldCheck)
{
@@ -500,7 +506,7 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
}
else
{
ulong r3 = ra.Q0M3 + (ra.Step3 * (idx % 3UL));
ulong r3 = ra.Q0M3 + (ra.Step3 * idxMod3);
if (r3 >= 6UL) r3 -= 6UL;
if (r3 >= 3UL) r3 -= 3UL;
if (r3 == 0UL)
@@ -509,7 +515,7 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
}
else
{
ulong r5 = ra.Q0M5 + (ra.Step5 * (idx % 5UL));
ulong r5 = ra.Q0M5 + (ra.Step5 * idxMod5);
if (r5 >= 15UL) r5 -= 15UL;
if (r5 >= 10UL) r5 -= 10UL;
if (r5 >= 5UL) r5 -= 5UL;
@@ -573,7 +579,9 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
ResidueAutomatonArgs ra, ArrayView<int> found, ArrayView1D<uint, Stride1D.Dense> smallCycles)
{
ulong idx = (ulong)index.X;
ulong r10 = ra.Q0M10 + (ra.Step10 * idx).Mod10(); r10 -= (r10 >= 10UL) ? 10UL : 0UL;
ulong idxMod3 = idx.Mod3();
ulong idxMod5 = idx.Mod5();
ulong r10 = ra.Q0M10 + (ra.Step10 * idx).Mod10(); r10 -= (r10 >= 10UL) ? 10UL : 0UL;
bool shouldCheck = r10 != 5UL;
if (shouldCheck)
{
@@ -584,7 +592,7 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
}
else
{
ulong r3 = ra.Q0M3 + (ra.Step3 * (idx % 3UL));
ulong r3 = ra.Q0M3 + (ra.Step3 * idxMod3);
if (r3 >= 6UL) r3 -= 6UL;
if (r3 >= 3UL) r3 -= 3UL;
if (r3 == 0UL)
@@ -593,7 +601,7 @@ public ref struct GpuKernelLease(IDisposable limiter, GpuContextLease gpu, Kerne
}
else
{
ulong r5 = ra.Q0M5 + (ra.Step5 * (idx % 5UL));
ulong r5 = ra.Q0M5 + (ra.Step5 * idxMod5);
if (r5 >= 15UL) r5 -= 15UL;
if (r5 >= 10UL) r5 -= 10UL;
if (r5 >= 5UL) r5 -= 5UL;
PerfectNumbers.Core/Gpu/MersenneNumberIncrementalGpuTester.cs
+2 -2
View File
@@ -25,8 +25,8 @@ public class MersenneNumberIncrementalGpuTester(GpuKernelType kernelType, bool u
var incKernel = gpuLease.IncrementalKernel;
ulong step10 = (exponent.Mod10() << 1).Mod10();
ulong step8 = ((exponent & 7UL) << 1) & 7UL;
ulong step3 = ((exponent % 3UL) << 1) % 3UL;
ulong step5 = ((exponent % 5UL) << 1) % 5UL;
ulong step3 = (exponent.Mod3() << 1).Mod3();
ulong step5 = (exponent.Mod5() << 1).Mod5();
GpuUInt128 twoPGpu = (GpuUInt128)twoP;
var smallCyclesView = GpuKernelPool.EnsureSmallCyclesOnDevice(accelerator);
ResiduePrimeViews primeViews = default;
PerfectNumbers.Core/Gpu/MersenneNumberLucasLehmerGpuTester.cs
+1 -1
View File
@@ -51,7 +51,7 @@ public class MersenneNumberLucasLehmerGpuTester
// Early rejections aligned with incremental/order sieves, but safe for small p:
// - If 3 | p and p != 3, then 7 | M_p -> composite.
// - If p ≡ 1 (mod 4) and p shares a factor with (p-1), reject fast.
if (exponent % 3UL == 0UL && exponent != 3UL)
if (exponent.Mod3() == 0UL && exponent != 3UL)
{
return false;
}
PerfectNumbers.Core/Gpu/MersenneNumberOrderGpuTester.cs
+4 -4
View File
@@ -27,10 +27,10 @@ public class MersenneNumberOrderGpuTester(GpuKernelType kernelType, bool useGpuO
var foundBuffer = accelerator.Allocate1D<int>(1);
UInt128 exponent128 = exponent;
exponent128.Mod10_8_5_3(out ulong exponentMod10, out ulong exponentMod8, out ulong exponentMod5, out ulong exponentMod3);
ulong step10 = (exponentMod10 << 1) % 10UL;
ulong step8 = (exponentMod8 << 1) & 7UL;
ulong step3 = (exponentMod3 << 1) % 3UL;
ulong step5 = (exponentMod5 << 1) % 5UL;
ulong step10 = (exponentMod10 << 1).Mod10();
ulong step8 = (exponentMod8 << 1) & 7UL;
ulong step3 = (exponentMod3 << 1).Mod3();
ulong step5 = (exponentMod5 << 1).Mod5();
try
{
UInt128 remaining;