cahlen/flint-hills-cuda

Flint Hills Series Partial Sums

Computes partial sums S_N = sum_{n=1}^N 1/(n^3 sin^2(n)) using quad-double spike arithmetic + Kahan-compensated bulk.

Usage

import torch
from kernels import get_kernel

kernel = get_kernel("cahlen/flint-hills-cuda")
result = flint_hills.partial_sum(max_n=1000000000)

Compile (standalone)

nvcc -O3 -arch=sm_90 -o flint_hills flint_hills/flint_hills.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 Not peer-reviewed.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.