UC San Diego Part of National Hub for Large-scale Neuromorphic Computing

"We are excited to launch THOR as an open platform for large-scale neuromorphic computing.  Broadly accessible to the research communities in computational neuroscience and artificial intelligence, THOR will give the user direct access to a range of powerful neuromorphic hardware modeling and emulating brain function at unprecedented scale,” said Cauwenberghs, a professor in the UC San Diego Shu Chien-Gene Lay Department of Bioengineering and director of the UC San Diego Institute for Neural Computation. 

“We see tremendous potential in leveraging this community resource to democratize the design and application of neuromorphic computing, lowering entry barriers that have traditionally limited access to experts and developers of the technology, thereby accelerating widespread adoption and shared developments while engaging the broader community in research, education, and training at the crossroads between multiple AI-centered STEM disciplines,” he added.

The THOR infrastructure will enable richer understanding of computational models, algorithms, neuromorphic hardware, and engineered test cases, supporting research in neuroscience and a wide range of application domains that benefit from bio-inspired processing, the researchers said. 

"We plan to design a national hub for open access large scale neuromorphic platforms, through close-knit industry partnerships," said Dhireesha Kudithipudi, principal investigator at the University of Texas at San Antonio. "The field is at a pivotal moment and ensuring access to a broader group of researchers is critical at this stage. This initiative reflects a community-driven approach, shaping a framework designed by and for the community."

The THOR infrastructure provides :

• Cutting-edge open-source software frameworks, libraries and tools.
• Collaborative community-driven workshops and knowledge sharing.
• Competitive challenges to foster innovation and collective advancement.
• Comprehensive standardized benchmarks for rigorous platform evaluation.

"By making bio-inspired computing resources available to a wider community of researchers in computer science, neuroscience, and computational physics, this project will contribute to democratizing access to advanced tools and fostering breakthroughs in energy-efficient, resilient AI through neuromorphic computing,” said NSF Program Director Andrey Kanaev. 

Additionally, the THOR team will develop training and educational materials that will cover the fundamentals of neuromorphic learning algorithms and systems, and all resources will be available through open-platforms to researchers, facilitating integration into both undergraduate and graduate curricula.

Related story: Funding provides researchers nationally with open access to large-scale, state-of-the-art neuromorphic computing platforms

Learn more about research and education at UC San Diego in: Artificial Intelligence