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

Refine Montgomery GPU helpers to avoid kernel branching

Browse Source
This commit is contained in:
Marek Lemańczyk committed 2025-10-22 01:43:10 +02:00
1 parent f12919e220
commit 24eaec7a21
5 files changed
+46 -22
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

PerfectNumbers.Core/Gpu/PrimeOrderGpuHeuristics.cs
+1 -1
View File
@@ -652,7 +652,7 @@ internal static partial class PrimeOrderGpuHeuristics
return 0UL;
}
return ULongExtensions.Pow2MontgomeryModWindowedGpu(divisorData, exponent, keepMontgomery: false);
return ULongExtensions.Pow2MontgomeryModWindowedGpuConvertToStandard(divisorData, exponent);
}
private static Action<AcceleratorStream, Index1D, ArrayView1D<ulong, Stride1D.Dense>, MontgomeryDivisorData, ArrayView1D<ulong, Stride1D.Dense>> GetPow2ModKernel(Accelerator accelerator)