5 papers
Architecture, Dataflow and Physical Design Implications of 3D-ICs for DNN-Accelerators
Jan Moritz Joseph, Ananda Samajdar, Lingjun Zhu +4
The everlasting demand for higher computing power for deep neural networks (DNNs) drives the development of parallel computing architectures. 3D integration, in which chips are int…
CLAN: Continuous Learning using Asynchronous Neuroevolution on Commodity Edge Devices
Parth Mannan, Ananda Samajdar, Tushar Krishna
Recent advancements in machine learning algorithms, especially the development of Deep Neural Networks (DNNs) have transformed the landscape of Artificial Intelligence (AI). With e…
SCALE-Sim: Systolic CNN Accelerator Simulator
Ananda Samajdar, Yuhao Zhu, Paul Whatmough +2
Systolic Arrays are one of the most popular compute substrates within Deep Learning accelerators today, as they provide extremely high efficiency for running dense matrix multiplic…
GeneSys: Enabling Continuous Learning through Neural Network Evolution in Hardware
Ananda Samajdar, Parth Mannan, Kartikay Garg +1
Modern deep learning systems rely on (a) a hand-tuned neural network topology, (b) massive amounts of labeled training data, and (c) extensive training over large-scale compute res…
Euphrates: Algorithm-SoC Co-Design for Low-Power Mobile Continuous Vision
Yuhao Zhu, Anand Samajdar, Matthew Mattina +1
Continuous computer vision (CV) tasks increasingly rely on convolutional neural networks (CNN). However, CNNs have massive compute demands that far exceed the performance and energ…