Recently, we announced that Cerebras will support the U.S. National Science Foundation’s (NSF) National Artificial Intelligence Research Resource (NAIRR) Pilot with a remote access grant of up to 4 exaFLOPs of AI compute for large language and generative AI model training using CS-2 systems. This unprecedented compute capacity will enable NAIRR Pilot researchers to train AI models in days on Cerebras vs. months on traditional GPU clusters, significantly accelerating their ability to probe the frontiers of science and engineering in AI in the national interest.
AI can fuel innovation in scientific discovery and economic growth; however, the frontiers of this emerging technology rely on access to computational power and data. To realize the transformative potential of AI, the White House, National Science Foundation, and Congress are developing the NAIRR as a widely accessible AI research cyberinfrastructure to democratize the AI research landscape in the U.S. The NAIRR Pilot is a shared initiative providing cutting edge AI infrastructure and tools to NSF, participating researchers, and students to inform future NAIRR design, expand access, and enhance collaborative research on AI to maintain and extend U.S. technology leadership in AI for science and engineering. By democratizing AI research infrastructure, NAIRR not only fosters innovation but also aims to bridge the gap between emerging AI technologies and their practical applications. NAIRR’s role in the U.S. is vital, as it not only advances AI technology but also ensures that the benefits of AI research are broadly distributed across various sectors, enhancing the country’s competitive edge in technology.
NAIRR’s national research cyberinfrastructure will support U.S. AI researchers to advance safe, secure, and trustworthy AI through providing compute resources in addition to data software, testbeds, and educational training. Furthermore, the pilot’s operations will be organized into four focus areas:
- NAIRR Open will enable open AI research through access to diverse AI resources via the NAIRR Pilot Portal and coordinated allocations.
- NAIRR Secure, co-led by NIH and DOE, will enable AI research requiring privacy and security-preserving resources and will assemble exemplary privacy-preserving resources.
- NAIRR Software will facilitate and investigate the inter-operable use of AI software, platforms, tools, and services for NAIRR pilot resources.
- NAIRR Classroom will reach new communities through education, training, user support, and outreach.
Cerebras is proud to participate and support the NAIRR Pilot. Cerebras will support the NAIRR Open focus area by offering access to our leading AI supercomputers as well as our world-class expertise in model development. These resources will not only empower researchers but also facilitate breakthroughs in AI that can be applied in diverse fields.
According to Dr. Andy Hock, Cerebras’ SVP, “The Cerebras team is thrilled to support the NAIRR Pilot to help build a national AI research infrastructure that will expand access to world-class AI compute and radically accelerate scientific AI research – program goals that are central to our company’s mission, as well. By contributing access to exaFLOPs of AI supercomputing power and support from our expert ML/AI engineering teams, we aim to help Pilot users accelerate and scale their work, enable NAIRR success, and meaningfully advance our nation’s leadership in AI computing and research.”
Supporting the NAIRR Pilot is just one facet of our extensive efforts to accelerate national interest in AI. By providing exceptional AI computing resources and expertise, we not only showcase the capabilities of our systems for large-scale AI research but also reinforce our status as a leader in U.S. AI technology.
Researchers can learn more about the NAIRR Pilot from the press release here: https://new.nsf.gov/news/democratizing-future-ai-rd-nsf-launch-national-ai
Related Posts
August 28, 2024
Integrating LLMs and Software Engineering for Self-Refining Copy Creation
Discover how to build an AI agent that generates marketing copy efficiently…
August 28, 2024
ReadAgent: Bringing Gist Memory to AI
Learn how gist memory improves long context handling for large language models.…