bfloat16

oneAPI Deep Neural Network Library (oneDNN)

Library for specialized dense and sparse matrix operations, and deep learning primitives.

Half-precision floating point types f16 and bf16 for Rust.

Julia implementation for the BFloat16 number type

IHP 130nm ASIC tapeout of a 2x2 bfloat16 matrix matrix multiplication with DFT infrastructure. Iteration on the previous accelerator taped out on GF180.

Round matrix elements to lower precision in MATLAB

C++ template library for floating point operations

RTL implementation of a performance/area optimized bfloat16 adder and multiplication.

Floating-Point Arithmetic Library for Z80

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

IEEE 754-style floating-point converter

A LLaMA2-7b chatbot with memory running on CPU, and optimized using smooth quantization, 4-bit quantization or Intel® Extension For PyTorch with bfloat16.

Pushing the bf16 multiplication clock frequency to the max on the nominal corner on IHP 130nm 5L node.

A JAX implementation of stochastic addition.

Bfloat16 conversion utilities for Go/Golang

Customizable floating point types, with all standard floating point operations implemented from scratch.

100Mbps Ethernet accelerator wrapper.

A Pytorch implementation of stochastic addition.

A lightweight C++ implementation of the Brain Floating Point (bfloat16) format.

Basic linear algebra routines implemented using the chop rounding function