hyperfold

Native high-performance linear algebra for Python.

Hyperfold combines a modern C++23 computing core with a NumPy-first API, explicit memory behaviour, and native CPU and CUDA execution.

PYPython API for NumPy, CuPy and PyTorch
APIZero-copy interchange with CuPy and PyTorch
SIMDExtensible CPU kernels for AVX2, AVX-512, ARM NEON and Apple Silicon
CUDAExtensible CUDA kernels with native support through NVIDIA Blackwell
LINALGDense and structured linear algebra
SOLVERSolver and factorisation pipelines

Documentation

import cupy as cp
import torch
import hyperfold as hf

H = [[4, 1-1j], [1+1j, 3]]  # Hermitian
A = cp.asarray(H, dtype=cp.complex64)
b = cp.asarray([1+2j, 3-1j], A.dtype)

lu = hf.lu_factor(A)       # factor once
x1 = lu.solve(b)
x2 = lu.solve(2 * b)       # solve many

t = torch.ones(8, device="cuda")
y = hf.scale(t, 2)          # zero-copy

braketware

hyperfold