 mlemanczyk
|
74837a407e
|
MersenneDivisorCycles converted to device template.
|
2025-12-19 04:03:27 +01:00 |
|
|
mlemanczyk
|
3c80221105
|
MontgomeryDivisorDataPool.Rent / .Return removed as we no longer pool Montgomery divisor data. UInt128Extensions.MontgomeryDivisorDataPool variables optimized.
|
2025-12-16 02:27:29 +01:00 |
|
|
mlemanczyk
|
05d3061c47
|
MethodImplOptions.AggressiveOptimization & MethodImplOptions.AggressiveInlining optimized. Small optimizations.
|
2025-12-10 19:59:51 +01:00 |
|
|
mlemanczyk
|
941733905b
|
Hybrid device mode implemented as the current --mersenne=bydivisor CPU execution path. CPU path is now limited to CPU calculations, only.
|
2025-12-01 03:45:39 +01:00 |
|
|
mlemanczyk
|
c311b27dcc
|
Prime testing using divisor tables depending on the last digit separated to its own class. PrimeTester switched back to standard endings.
|
2025-11-27 16:45:02 +01:00 |
|
|
mlemanczyk
|
13498dc460
|
Most device buffers integrated into PrimeOrderCalculatorAccelerator. PrimeOrderCalculator switched to partial factoring on GPU.
|
2025-11-26 03:16:39 +01:00 |
|
|
mlemanczyk
|
47521e327a
|
Cleanup
|
2025-11-25 00:50:18 +01:00 |
|
|
mlemanczyk
|
91894637af
|
Aggressive inlining added to simple switch methods.
|
2025-11-24 22:54:04 +01:00 |
|
|
mlemanczyk
|
b9ca95bf8c
|
PrimeOrderCalculatorAccelerator.Rent is now created outside of the loops and re-used for all calculations.
|
2025-11-24 22:51:53 +01:00 |
|
|
mlemanczyk
|
070e2a14fd
|
MontgomeryDivisorData converted to a pooled class to limit GC pressure. MontgomeryDivisorDataGpu added for GPU operations. PrimeOrderCalculator kernels precompiled during warmup.
|
2025-11-24 15:14:31 +01:00 |
|
|
mlemanczyk
|
497b1fcf89
|
Kernels replaced with pre-compiled actions for better performance
|
2025-11-20 03:00:11 +01:00 |
|
|
mlemanczyk
|
df04ba601c
|
Various small optimizations and cleanups. Methods made static, where possible. Duplicated & slow PrimeTester.BinaryGcd implementation removed. Optimized version was used instead.
|
2025-11-20 01:43:29 +01:00 |
|
|
mlemanczyk
|
5cac51479e
|
GPU memory allocations protected with lock to prevent shared memory access issues. GPU usage limited for the best performance.
|
2025-11-19 16:49:01 +01:00 |
|
|
mlemanczyk
|
66ca969929
|
Naming improvements and small cleanup
|
2025-11-18 03:43:52 +01:00 |
|
|
mlemanczyk
|
bf7c4bab39
|
Accelerator buffers integrated
|
2025-11-18 01:13:12 +01:00 |
|
|
mlemanczyk
|
f630bfb598
|
HeuristicCombinedPrimeTesterAccelerator integrated with PrimeOrderCalculatorAccelerator.
|
2025-11-17 17:38:23 +01:00 |
|
|
mlemanczyk
|
183b147135
|
Heuristic tables split into 2 groups
|
2025-11-17 16:04:27 +01:00 |
|
|
mlemanczyk
|
be435b7760
|
Kernels made part of the accelerators for better performance.
|
2025-11-13 18:52:46 +01:00 |
|
|
mlemanczyk
|
0d2a88c45f
|
GpuPrimeWorkLimiter.Acquire() / .Release() removed to save time on the hot-path with billions of calls. We'll rely on AcceleratorStreamPool limits and synchronization.
|
2025-11-13 17:52:13 +01:00 |
|
|
mlemanczyk
|
5bc65464d0
|
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.
|
2025-11-12 21:18:30 +01:00 |
|
|
mlemanczyk
|
0de9871529
|
Functionality and classes separated. Kernel initialization made dependent on a flag. Cleanup.
|
2025-11-10 18:35:50 +01:00 |
|
|
mlemanczyk
|
d0b1639a5d
|
GPU stabilization improvements to prevent calculation crashes
|
2025-11-09 21:52:16 +01:00 |
|
|
mlemanczyk
|
de2a78b27b
|
Streams added to all kernels and GPU operations to separate threads from each other.
|
2025-11-06 18:53:55 +01:00 |
|
|
mlemanczyk
|
64d615c402
|
Various optimizations to separate non-static data between threads and resolve read/write access issue.
|
2025-11-04 20:45:56 +01:00 |
|
|
Marek Lemańczyk
|
4c4d4243fb
|
Merge remote-tracking branch 'github/codex/optimize-evenperfectbitscanner-for-static-arrays-khd9w8' into work
# Conflicts:
# PerfectNumbers.Core/Gpu/GpuContextPool.cs
# PerfectNumbers.Core/Gpu/GpuKernelLease.cs
# PerfectNumbers.Core/Gpu/GpuKernelPool.cs
# PerfectNumbers.Core/Gpu/GpuStaticTableInitializer.cs
# PerfectNumbers.Core/Gpu/Kernels/PrimeOrderSpecialMaxKernels.cs
# PerfectNumbers.Core/PrimeOrderCalculator.Cpu.cs
# PerfectNumbers.Core/PrimeOrderCalculator.Gpu.cs
# PerfectNumbers.Core/PrimeTester.cs
|
2025-11-03 23:59:12 +01:00 |
|
|
Marek Lemańczyk
|
76dfe108a9
|
Address feedback on GPU pooling and heuristic factoring
|
2025-11-02 11:22:49 +01:00 |
|
|
Marek Lemańczyk
|
ab6a416de2
|
Warm GPU resources before thread startup
|
2025-11-02 11:22:28 +01:00 |
|
|
mlemanczyk
|
2281d9535e
|
Revert "Merge pull request #251 from mlemanczyk/codex/refactor-prime-checking-on-gpu-n5t2pm"
This reverts commit 936ce239db, reversing
changes made to e14b77e504.
|
2025-11-02 11:22:14 +01:00 |
|
|
Marek Lemańczyk
|
8c662c5c85
|
Address feedback on GPU pooling and heuristic factoring
|
2025-10-31 21:47:42 +01:00 |
|
|
Marek Lemańczyk
|
6864b316d5
|
Warm GPU resources before thread startup
|
2025-10-31 16:56:25 +01:00 |
|
|
mlemanczyk
|
e14b77e504
|
Batch size made dynamic - dependent on the value
|
2025-10-30 02:43:15 +01:00 |
|
|
Marek Lemańczyk
|
f9504a6a19
|
Use atomic flag writes in heuristic trial division
|
2025-10-29 20:40:40 +01:00 |
|
|
Marek Lemańczyk
|
f7a0627ce5
|
Align heuristic prime testers with squared limits
|
2025-10-29 19:35:59 +01:00 |
|
|
Marek Lemańczyk
|
a4ac82e74d
|
Inline GPU kernel state into prime tester leases
|
2025-10-29 18:13:51 +01:00 |
|
|
Marek Lemańczyk
|
aab785650d
|
Manage GPU lease buffers without kernel scratch pool
|
2025-10-29 16:53:22 +01:00 |
|
|
Marek Lemańczyk
|
13b59abb1c
|
Embed GPU scratch buffers in context leases
|
2025-10-29 16:25:38 +01:00 |
|
|
Marek Lemańczyk
|
b753654ebb
|
Enforce valid divisor selectors in heuristic combined tester
|
2025-10-29 15:32:40 +01:00 |
|
|
Marek Lemańczyk
|
b0158a844a
|
Use divisor squares to bound heuristic combined trial division
|
2025-10-29 15:13:54 +01:00 |
|
|
mlemanczyk
|
b2865b54ce
|
Tiny optimization
|
2025-10-29 14:04:25 +01:00 |
|
|
mlemanczyk
|
4c79c4087e
|
Various optimizations
|
2025-10-29 14:00:38 +01:00 |
|
|
Marek Lemańczyk
|
f49963c4f6
|
Restore Math.Sqrt heuristic bound
|
2025-10-29 01:43:12 +01:00 |
|
|
Marek Lemańczyk
|
5a9e103bd5
|
Replace heuristic sqrt with log2-based estimate
|
2025-10-29 01:29:12 +01:00 |
|
|
Marek Lemańczyk
|
eeeff42be6
|
Replace heuristic sqrt with integer algorithm
|
2025-10-29 01:02:31 +01:00 |
|
|
Marek Lemańczyk
|
b54f4b6bd6
|
Simplify combined heuristic divisor selection
|
2025-10-29 00:44:46 +01:00 |
|
|
Marek Lemańczyk
|
b2cb17a96f
|
Return to one-A-one-B combined schedule
|
2025-10-29 00:44:39 +01:00 |
|
|
Marek Lemańczyk
|
f993561cd5
|
Eagerly build combined divisors and restore two-A-one-B default
|
2025-10-28 23:43:15 +01:00 |
|
|
Marek Lemańczyk
|
e8d163f595
|
Add three-A-two-B combined divisor schedule
|
2025-10-28 23:22:16 +01:00 |
|
|
Marek Lemańczyk
|
465b40bf35
|
Add three-A-one-B schedule to combined tester
|
2025-10-28 23:09:41 +01:00 |
|
|
Marek Lemańczyk
|
4c445f2a42
|
Switch combined heuristic divisors to 1A-1B schedule
|
2025-10-28 22:55:17 +01:00 |
|
|
Marek Lemańczyk
|
41bff698e8
|
Add combined heuristic prime tester with shared divisor arrays
|
2025-10-28 22:40:51 +01:00 |
|