Initial PollardRho implementation for GPU. Added pool for accelerator streams shared across the threads. GPU processing performance and stabilization improvements. GpuUInt128.BinaryGcd switched to the version which outperformed other in benchmarks.

PerfectNumbers.Core/Gpu/Kernels/PrimeOrderKernels.cs

@@ -2,6 +2,7 @@ using System.Numerics;
 using System.Runtime.CompilerServices;
 using ILGPU;
 using ILGPU.Runtime;
+using PerfectNumbers.Core.Cpu;
 
 namespace PerfectNumbers.Core.Gpu;
 
@@ -1410,5 +1411,4 @@ internal static partial class PrimeOrderGpuHeuristics
 		ulong directMask = (1UL << windowBitCount) - 1UL;
 		return exponent.Low & directMask;
 	}
-
 }