Search
  • Salman Avestimehr

Coded Computing: An Overview


Coded Computing is a transformative framework that we have been developing over the past few years. It combines coding theory with distributed computing to inject computation redundancy in an unorthodox coded form in order to alleviate two major performance bottlenecks that arise in large-scale distributed computing/machine learning:

  1. reducing the computation latency by enabling the completion of the distributed task without waiting for slower nodes to finish computing;

  2. reducing communication requirements by providing coding opportunities during data exchange.


Coded computing can also enable strong (information-theoretic) security and private in distributed learning/computing.


Here is a short overview of our recent results in coded computing.


162 views

University of Southern California

© 2019 by Salman Avestimehr