An Open-Source Library for Low-Power Approximate Computing Modules
The “lpACLib” library contains the VHDL description of accurate and approximate versions of several arithmetic modules (like adders and multiplier of different bit-widths) and accelerators. Moreover, it also provides the corresponding software behavioral models/implementations developed in C and MATLAB to enable quality characterization. Besides our novel designs, it also contains implementations for several state-of-the-art arithmetic modules and their approximate versions. This open-source library facilitates research and development in approximate computing at higher abstraction levels, and to facilitate reproducible research and comparisons. In case of usage, please refer to our publication: Muhammad Shafique, Rehan Hafiz, Semeen Rehman, Walaa El-Harouni, Jörg Henkel, "Cross-Layer Approximate Computing: From Logic to Architectures", Design Automation Conference (DAC), 2016. Contributors: Authors, Vanshika Baoni, M. Abdullah Hanif http://ces.itec.kit.edu/lpACLib.php
Library of Approximate Adders
We provide MATLAB and Verilog Models of GeAr, and previously proposed adders (ACA-I, ETAII, ACA-II and GDA) at http://sourceforge.net/projects/approxadderlib/ GeAr is a low latency Generic Accuracy Configurable Adder that provides a higher number of potential configurations compared to state-of-the-art approximate adders, thus enabling a high degree of flexibility and trade-off between performance and output quality. These MATALB and Verilog models can allow software programmer as well as hardware designers to evaluate their code and design. To the best of our knowledge, this is the first open-source library of approximate adders that facilitates reproducible comparisons and further research and development in this direction across various layers of design abstraction. This work is a result of collaborative effort between Chair for Embedded Systems (CES) at Karlsruhe Institute of Technology (KIT), Germany and Vision Image and Signal Processing (VISpro) Lab at SEECS-NUST, Pakistan.
These classes are useful for signal processing in Matlab or C++. They bring together tools and methods which may be used interchangeably for Matlab and C++. Their initial use is in conjunction with work towards my degree at UC Berkeley.
Magnetostatic finite differences simulation tool based on GNU Octave and a Perl/Tk user interface.