cahlen/zaremba-density-cuda

Zaremba Density Enumeration

Enumerates all CF denominators <= N with partial quotients from digit set A using persistent-thread DFS + bitset marking.

Usage

import torch
from kernels import get_kernel

kernel = get_kernel("cahlen/zaremba-density-cuda")
result = zaremba.count_representable(max_d=1000000, digits=torch.tensor([1,2,3]))

Compile (standalone)

nvcc -O3 -arch=sm_90 -o zaremba_density zaremba_density/zaremba_density_gpu.cu -lm

Results

All computation results are open:

• Website: bigcompute.science
• Datasets: huggingface.co/cahlen
• Source: github.com/cahlen/idontknow

Citation

@misc{humphreys2026bigcompute,
  author = {Humphreys, Cahlen},
  title = {bigcompute.science: GPU-Accelerated Computational Mathematics},
  year = {2026},
  url = {https://bigcompute.science}
}

Human-AI collaborative. Not peer-reviewed. All code and data open.