CUDA-based solver for PerimeterX PoW challenges. Massively faster than CPU solutions.
GPU-accelerated solver for PerimeterX PoW challenges. Made for high-volume applications where CPU solvers can't keep up.
Benchmark results using a test PoW string with difficulty 20:
"I0I0I0|1|5262cb5ce06ac910131713c159b214c89f67a451c547434981bc6aaf812|dcaf6a6c99a78dd46ebe06ca177abdf5b19accf9647b567fd4da736f4676f67d|20|false"
| Method | Total Time (250 solves) | Avg Time Per Solve | Performance Gain |
|---|---|---|---|
| CPU | 78,889 ms | 315.56 ms | Baseline |
| GPU | 428 ms | 1.71 ms | ~184x faster |
Testing performed using C# application calling the solver via DLL integration
- CUDA-based implementation for maximum GPU utilization
- Compatible with PerimeterX PoW challenge format
- C# interop through DLL interface
- Handles high-difficulty challenges efficiently
- CUDA-compatible NVIDIA GPU
- CUDA Toolkit
- Compatible operating system (Windows/Linux)
- Better batching for higher throughput
- Fix memory handling for proper multi-threading
- Multi-GPU support
- Command-line interface
nvcc --shared -o CudaPowSolver.dll powsolver.cuUse at your own risk. Make sure your usage complies with PerimeterX's terms.