In-Memory Nearest Neighbor Search with FeFET Multi-Bit Content-Addressable Memories
Arman Kazemi1,a , Mohammad Mehdi Sharifi1,b, Ann Franchesca Laguna1, Franz Müllery2, Ramin Rajaei1, Ricardo Olivoy2, Thomas Kämpfey2, Michael Niemier1 and X. Sharon Hu1
1University of Notre Dame, IN, USA
aakazemi@nd.edu
bmsharif1@nd.edu
2Fraunhofer IPMS-CNT, Dresden, Germany
ABSTRACT
Nearest neighbor (NN) search is an essential operation in many applications, such as one/few-shot learning and image classification. As such, fast and low-energy hardware support for accurate NN search is highly desirable. Ternary contentaddressable memories (TCAMs) have been proposed to accelerate NN search for few-shot learning tasks by implementing L1 and Hamming distance metrics, but they cannot achieve softwarecomparable accuracies. This paper proposes a novel distance function that can be natively evaluated with multi-bit contentaddressable memories (MCAMs) based on ferroelectric FETs (Fe- FETs) to perform a single-step, in-memory NN search. Moreover, this approach achieves accuracies comparable to floating-point precision implementations in software for NN classification and one/few-shot learning tasks. As an example, the proposed method achieves a 98.34% accuracy for a 5-way, 5-shot classification task for the Omniglot dataset (only 0.8% lower than softwarebased implementations) with a 3-bit MCAM. This represents a 13% accuracy improvement over state-of-the-art TCAM-based implementations at iso-energy and iso-delay. The presented distance function is resilient to the effects of FeFET device-to-device variations. Furthermore, this work experimentally demonstrates a 2-bit implementation of FeFET MCAM using AND arrays from GLOBALFOUNDRIES to further validate proof of concept.
Keywords: Nearest Neighbor Search, Content-Addressable Memory, Multi-Bit Design, Ferroelectric FET.
