Cultivating Multidisciplinary Research and Education on GPU Infrastructure for Mid-South Institutions at the University of Memphis: Practice and Challenge

2025-04-21

AI Summary

Resource-constrained institutions in the U.S. South-Central region lack access to high-performance computing (HPC) infrastructure for AI and data science education and research. Method: This project designed, deployed, and operates iTiger—the region’s first mid-scale shared GPU cluster—under a novel, sustainable operational model integrating seed funding, curriculum-embedded instruction, hands-on workshops, and community outreach. The system integrates an HPC job scheduler, AI-focused pedagogical platform, interdisciplinary course framework, and comprehensive user support infrastructure to tightly couple computational resources with teaching and research missions. Contribution/Results: The initiative significantly increased HPC utilization and AI-related scholarly output across multiple regional institutions. It enabled applied interdisciplinary projects in precision agriculture, intelligent transportation, and health informatics. Furthermore, it established a scalable, replicable blueprint for equitable, edge-region HPC capacity building—demonstrating how shared GPU infrastructure can advance both educational equity and research competitiveness in underserved academic communities.

Technology Category

• Humans and AI: Other Foundations of Human Computation & AI
• Machine Learning: Hardware-aware ML
• Data Mining & Knowledge Management: Scalability, Parallel & Distributed Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

To support rapid scientific advancement and promote access to large-scale computing resources for under-resourced institutions at the Mid-South region, the University of Memphis (UofM) established the first regional mid-scale GPU cluster, iTiger, a valuable high-performance computing (HPC) infrastructure. In this study, we present our continuous efforts to manage the critical cyberinfrastructure and provide essential computing supports for educators, students, and researchers in AI, data sciences, and related scientific fields in the Mid-South region, such as precision agriculture, smart transportation, and health informatics. We outline our initiatives to broaden CI adoptions across regional computing-related scientific and engineering fields, such as seed grant, workshop trainings, course integration, and other outreach activities. While we've observed promising outcomes of regional CI adoptions, we will discuss insights and challenges of Mid-South CI users, which can inspire other institutions to implement similar programs.

Problem

• Providing GPU-based HPC resources for under-resourced Mid-South institutions
• Supporting multidisciplinary research in AI, data science, and regional applications
• Expanding cyberinfrastructure adoption through education and outreach initiatives

Innovation

• Established regional mid-scale GPU cluster iTiger
• Provided HPC supports for AI and data sciences
• Broadened CI adoptions via workshops and grants