*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Context based search using Content Addressable Memory
Committee:
Dr. Naeemi, Advisor
Dr. Mukhopadhyay, Chair
Dr. Romberg
Abstract: The objective of the proposed research is to design unconventional circuits to take advantage of non- volatile memories like Spin Orbit Torque (SOT) and Magnetoelectric (ME) Magnetic random access memory (MRAM). Content addressable memories (CAMs) have proved themselves to be worthy contenders for data-intensive applications. CAMs have been used in many applications for performing similarity and nearest neighbor search. However, challenges like minimum detectable Hamming distance, discharge current saturation and energy and performance efficiency have limited their use. In recommendation models and hyperdimensional computing, the features/bits important to a query while searching through the dataset change for each search cycle based on ’context’. Here we will propose SOT and ME-based CAM design to perform NNS with ’context’. This allows us to improve energy efficiency of CAMs and improves the quality of search.