---

The San Diego Supercomputer Center (SDSC) at UC San Diego has been awarded a grant from the National Science Foundation (NSF) to enhance artificial intelligence (AI) research through the utilization of NVIDIA’s DGX Cloud platform. This initiative is part of the National AI Research Resource (NAIRR) Pilot program, which aims to provide U.S. researchers and educators with access to advanced AI resources

Leveling the Playing Field: The NAIRR Pilot Program

The NSF grant underscores the importance of equitable access to advanced computational resources. The NAIRR Pilot program, a federal initiative, aims to bridge the gap between well-funded institutions and smaller research centers by offering shared access to state-of-the-art AI platforms. This effort is crucial in ensuring that researchers, regardless of their institutional resources, can contribute to scientific breakthroughs in AI-driven fields like healthcare, climate science, and education.

NVIDIA DGX Cloud: The Engine of Innovation

Central to the San Diego Supercomputer Center’s (SDSC) collaboration with the National Science Foundation (NSF) is NVIDIA’s DGX Cloud platform, a revolutionary AI supercomputing infrastructure that transforms how researchers access and utilize computational resources. Unlike traditional on-premises hardware setups that require significant capital investment, physical space, and maintenance, NVIDIA DGX Cloud offers a fully managed, cloud-based solutiondesigned to handle the most computationally demanding tasks seamlessly.

Breaking Down the Key Features of NVIDIA DGX Cloud

1. 32-Node GPU Clusters for Unmatched Computational Power
   • At its core, DGX Cloud is built on high-performance Graphics Processing Units (GPUs), specialized processors ideal for parallel computing tasks common in AI and deep learning. The 32-node clusterconfiguration allows researchers to:
     • Run simulations, model training, or data analysis processes that would typically take weeks on standard systems, completing them in days or even hours.
     • Work on massive datasets required for fields like genomics, climate modeling, and advanced robotics.
     • Scale up their workloads without being constrained by local hardware limitations.
2. Generative AI and Deep Learning Capabilities
   • DGX Cloud excels at supporting generative AI—the technology behind tools like ChatGPT and DALL-E—and other advanced deep learning models.
   • Researchers can develop and refine Large Language Models (LLMs) or other neural networks used in cutting-edge AI applications, such as:
     • Text summarization, sentiment analysis, and chatbots.
     • Advanced image recognition for medical diagnostics.
     • Predictive analytics in fields like finance, agriculture, and disaster response.
   • The system provides pre-configured environments optimized for these AI tasks, reducing setup time and enabling faster experimentation.
3. A Streamlined Interface for Accessibility
   • Unlike traditional supercomputers, which often require significant expertise to configure and operate, DGX Cloud offers a user-friendly interface that democratizes access to advanced AI tools.
   • Researchers no longer need to invest in expensive physical infrastructure or hire dedicated IT staff to manage hardware. Instead, they can:
     • Access resources directly through their web browsers.
     • Focus on research and innovation rather than the technical complexities of system management.
     • Collaborate remotely with peers and share access to datasets or computational power.

Why DGX Cloud Matters to Researchers

Historically, access to such advanced computing power has been limited to elite institutions with substantial budgets. NVIDIA DGX Cloud removes these barriers, allowing researchers at smaller universities, rural institutions, or underfunded programs to leverage world-class AI tools. This democratization is a critical step toward enabling equitable participation in groundbreaking scientific research.

By integrating DGX Cloud into its resource offerings, SDSC provides researchers with an unparalleled platform to explore complex problems, develop transformative technologies, and push the boundaries of innovation. From speeding up medical breakthroughs to enabling real-time climate predictions, DGX Cloud empowers scientists to tackle challenges that were previously out of reach due to computational constraints.

This platform is more than a technical tool—it’s a catalyst for discovery, collaboration, and global impact.

Joe Monteiro, Chief Technology Officer at SDSC, stated:
“This grant allows us to bring the power of AI to a broader audience, ensuring that researchers nationwide have the tools they need to drive innovation.”

SDSC’s Expertise and Infrastructure

SDSC has long been at the forefront of computational innovation. Its flagship systems, such as the Expanse Supercomputer and the Voyager System, are already tailored to support complex AI tasks and deep learning applications. With the NSF grant, SDSC will:

• Develop the NAIRR Pilot portal, a centralized hub for researchers to access DGX Cloud and other advanced computing resources.
• Offer guidelines, training, and tools to optimize the use of NVIDIA’s platform.
• Combine its existing infrastructure with DGX Cloud, creating a seamless ecosystem for researchers.

Addressing Key Challenges in AI Research

The collaboration between SDSC and NVIDIA tackles several longstanding barriers in AI research:

1. Accessibility: Many institutions lack the infrastructure to support compute-heavy AI workloads, limiting their participation in cutting-edge research.
2. Scalability: NVIDIA’s DGX Cloud allows researchers to scale their experiments without needing costly on-premises systems.
3. Efficiency: SDSC’s tools and training programs will enable researchers to maximize the potential of DGX Cloud, reducing inefficiencies and accelerating results.

By addressing these challenges, the initiative empowers researchers to focus on discovery rather than logistical or computational hurdles.

Applications Across Diverse Fields

The NSF grant and NVIDIA resources will unlock new possibilities in AI research across various disciplines:

• Climate Science: AI models trained on DGX Cloud can analyze vast datasets to predict climate patterns and develop solutions for global challenges.
• Healthcare: From drug discovery to personalized medicine, the platform can accelerate innovations that improve patient outcomes.
• Disaster Response: AI can optimize resource allocation and predict the impact of natural disasters, helping communities prepare and recover more effectively.

The partnership between SDSC, NSF, and NVIDIA exemplifies the power of public-private collaboration in advancing technology. By combining federal support with cutting-edge industry expertise, this initiative sets a new standard for how resources can be shared to drive innovation. It also highlights the role of institutions like SDSC in bridging academia and industry to solve real-world problems.

Jensen Huang, CEO of NVIDIA, added:
“NVIDIA DGX Cloud is designed to unlock the potential of AI for all researchers, and we’re thrilled to see SDSC leading this charge through the NAIRR Pilot program.”

The Future of AI Research

Looking ahead, SDSC’s work with NVIDIA DGX Cloud will likely pave the way for:

• Multimodal AI Integration: Combining data from text, images, and other sources to create comprehensive models.
• Digital Twins: Developing AI-powered virtual representations of systems, such as cities or human organs, for simulation and optimization.
• AI Democratization: Expanding access to high-performance AI tools across underrepresented regions and institutions.

These advancements will not only benefit AI researchers but also catalyze breakthroughs in diverse fields that rely on advanced computation.

---