The International Conference for High Performance Computing, Networking, Storage, and Analysis
The International Conference for High Performance Computing, Networking, Storage, and Analysis
Authors: Craig Vineyard (Sandia National Laboratories), Suma Cardwell (Sandia National Laboratories), Catherine Schuman (University of Tennessee, Knoxville), Hector Gonzalez (SpiNNcloud), Ryad Benosman (Meta Platforms, Inc.), Robinson Pino (DOE Office of Advanced Scientific Computing Research)
Abstract: As AI and post-exascale demands reshape high performance computing, and neuromorphic systems push the boundaries of brain-inspired efficiency, a key question emerges: Does HPC need neuromorphic architectures to stay sustainable—or is neuromorphic computing reliant on HPC infrastructure to scale competitively?
This BoF session aims to unite researchers, practitioners, and industry leaders to explore overlapping interests between neuromorphic computing and HPC. The session will consist of a concise overview of the state of neuromorphic computing, insights from a diverse panel of experts, and an interactive discussion on mutual benefits and future collaboration.
Long Description: Neuromorphic computing is a brain-inspired approach to hardware and algorithm design. At a critical juncture it is posed to impact a range of fields such as HPC. As HPC increasingly integrates AI-enabled architectures and navigates the challenges of post-exascale computing, the intersection of these fields presents unique opportunities.
This BoF session aims to unite researchers, practitioners, and industry leaders to explore overlapping interests between neuromorphic computing and HPC. In particular, the BoF seeks to identify how neuromorphic can benefit HPC and what HPC needs from neuromorphic for adoption.
To pursue this goal, the session will begin with a short overview summarizing the state of neuromorphic computing. Many of the session leads are authors on the Nature 2025 manuscript titled "Neuromorphic computing at scale" [Kudithipudi D, Schuman C, Vineyard CM, Pandit T, Merkel C, Kubendran R, Aimone JB, Orchard G, Mayr C, Benosman R, Hays J. Neuromorphic computing at scale. Nature. 2025 Jan 23;637(8047):801-12] which provides a perspective on the neuromorphic landscape. Following this overview, four panelists will each present for five minutes, addressing key topics such as large-scale systems, the neuromorphic software stack, and perspectives from commercial and government sectors.
The second half of the BoF will focus on an interactive audience and panel discussion. Moderated by the Sandia Labs co-leaders, the audience and panel will be encouraged to discuss their perspective on challenges and needs from the traditional HPC and neuromorphic communities. To foster alignment with the HPC community, the discussion will emphasize topics relevant to both communities such as: the current state and availability of large-scale neuromorphic systems, parallels and challenges between programming large-scale neuromorphic versus traditional HPC paradigms, and the types of HPC-relevant applications that neuromorphic computing may be well-suited for.
To our knowledge, this BoF has not been held previously. Our intention is to create a collaborative environment that bridges the gap between the neuromorphic research community and traditional HPC practitioners. As part of this endeavor, we aspire to identify how neuromorphic computing relates to and differs from other paradigms the HPC field has engaged recently (e.g. GPUs, data-flow architectures, reconfigurable computation, and other specialized neural network accelerators). By fostering dialogue and collaboration, we aim to form a working group that will drive future research and development efforts at the intersection of these two fields.