half-precision 开源项目推荐（44 个）

stillwater-sc /universal

大量数制系统，为 AI、机器学习、计算机视觉、信号处理、CAE、EDA、控制、优化、估计和近似中的混合精度算法开发与优化提供自定义算术。

Maratyszcza /FP16

Conversion to/from half-precision floating point formats

vargalabs /h5cpp

C++17 templates between [stl::vector | armadillo | eigen3 | ublas | blitz++] and HDF5 datasets

petamoriken /float16

ES2025 float16 (IEEE 754 half-precision floating-point) ponyfill/proposal since 2017

x448 /float16

float16 provides IEEE 754 half-precision format (binary16) with correct conversions to/from float32

bindog /pytorch-model-parallel

A memory balanced and communication efficient FullyConnected layer with CrossEntropyLoss model parallel implementation in PyTorch

fengwang /float16_t

CPP20 implementation of a 16-bit floating-point type mimicking most of the IEEE 754 behavior. Single file and header-only.

artyom-beilis /float16

half float library for C and for z80

higham /chop

Round matrix elements to lower precision in MATLAB

shibatch /tlfloat

C++ template library for floating point operations

hma02 /cublasHgemm-P100

Code for testing the native float16 matrix multiplication performance on Tesla P100 and V100 GPU based on cublasHgemm

canbula /ieee754

Python module which finds the IEEE-754 representation of a floating point number.

KernelTuner /kernel_float

CUDA/HIP header-only library for low-precision (16 bit, 8 bit) and vectorized GPU kernel development

DW0RKiN /Floating-point-Library-for-Z80

Floating-Point Arithmetic Library for Z80

adelj88 /rocm_wmma_gemm

WMMA GEMM in ROCm for RDNA GPUs

minhhn2910 /cuda-half2

Convert CUDA programs from float data type to half or half2 with SIMDization

jizhuoran /caffe-android-opencl-fp16

Optimised Caffe with OpenCL supporting for less powerful devices such as mobile phones

enp1s0 /cuMpSGEMM

Fast SGEMM emulation on Tensor Cores

SomeRandomiOSDev /Half

Swift Half-Precision Floating Point

neslib /Neslib.Half

Half-Precision Floating-Point for Delphi