Beyond Floating Point - Posit C/C++ implementation

Open-Source Posit RISC-V Core with Quire Capability

c++ posit implementation

PositNN - Framework for training and inference with neural nets usings posits

posit (unum type III) integrated Tensorflow.

Posit multiplier Operator for the FloPoCo tool.

This system allows you to independently create and then combine the implementation of any format of real numbers and a set of benchmarks for testing the performance and correctness of calculations of the implementation in question. Posit (unum), float and bfloat tested.

QA Verification and Validation test suites for ALU and SFU hardware designs of number systems available in Universal

Universal Number Arithmetic. Posits with underflow support.

Modified version of the Support Vector Machines optimization problem for targeting low precision posit representation at inference time