Design Principles for Lifelong Learning AI Accelerators
RWTH Publications (RWTH Aachen)
Abstract
Lifelong learning - an agent's ability to learn throughout its lifetime - is a hallmark of biological learning systems and a central challenge for artificial intelligence (AI). The development of lifelong learning algorithms could lead to a range of novel AI applications, but this will also require the development of appropriate hardware accelerators, particularly if the models are to be deployed on edge platforms, which have strict size, weight, and power constraints. Here, we explore the design of lifelong learning AI accelerators that are intended for deployment in untethered environments. We identify key desirable capabilities for lifelong learning accelerators and highlight metrics to evaluate such accelerators. We then discuss current edge AI accelerators and explore the future design of lifelong learning accelerators, considering the role that different emerging technologies could play.
Authors 12
-
The University of Texas at San Antonio
Affiliation as printed
University of Texas at San Antonio , San Antonio , TX , USA
-
Affiliation as printed
Sandia National Laboratories , Albuquerque , NM , USA
-
Affiliation as printed
Argonne National Laboratory , Lemont , IL , USA
-
Rochester Institute of Technology
Affiliation as printed
Rochester Institute of Technology , Rochester , NY , USA
-
Emre Neftci Aachen
Forschungszentrum Jülich · RWTH Aachen University
Affiliation as printed
Forschungszentrum Jülich and RWTH Aachen , Aachen , Germany
-
Affiliation as printed
Tenstorrent Inc. , MA , USA
-
Affiliation as printed
University of Pisa , Pisa PI , Italy
-
United States Air Force Research Laboratory · Rome Laboratory
Affiliation as printed
Air Force Research Laboratory , Rome , NY , USA
-
Affiliation as printed
ECS Federal , Arlington , VA , USA
Cited by 1 stored of 1
1 result
No patents citing this paper on Lens.org (checked 2026-10-06).