Use Cases of Neuromorphic Co-Processors in Future HPC Environments
Authors: Catherine Schuman (Oak Ridge National Laboratory)
Abstract: With the looming end of Moore’s law and the end of Dennard scaling, the HPC community is exploring the use of specialized hardware as accelerators for certain tasks. Neuromorphic computing is a field in which neural networks are implemented in hardware to achieve intelligent computation with lower power and on a smaller footprint than traditional von Neumann architectures. Neuromorphic systems are compelling candidates for inclusion as co-processors in future HPCs, and they are suitable as co-processors for multiple types of applications. Here, we discuss neuromorphic systems as machine learning and graph algorithm accelerators. As more users are exposed to neuromorphic systems, we anticipate that even more use cases will arise.
Back to Women in HPC: Diversifying the HPC Community Archive Listing
Back to Full Workshop Archive Listing