UT Austin’s InfraAI Center for Systems Infrastructure and AI aims to tackle foundational challenges and drive innovation on these fronts. Run in close collaboration with industry partners, the InfraAI Center will focus on the co-evolution of AI and infrastructure systems, addressing today’s pressing challenges while reimagining the design of future systems. 

InfraAI’s research will be organized into two key pillars.

The first thrust will develop nimble AI stacks that natively embrace hardware and model heterogeneity, evolve with emerging AI workloads, deliver high performance and efficiency, and democratize access to AI. Today’s stacks are rigid—tightly coupling homogeneous hardware, models, and systems software—making them hard to adapt, inefficient, and accessible only to well-resourced organizations. InfraAI will re-architect training and inference stacks from the ground up, introducing abstractions that treat hardware diversity and AI model evolution as first-class concerns. These stacks will also embed learning algorithms designed to operate under constraints of compute, data, and connectivity—broadening access to AI infrastructure and democratizing innovation.

The second will develop intelligent infrastructures, embedding adaptive AI-driven decision-making in emerging cloud, edge, and consumer-device computing setups. Today’s infrastructures rely on manually designed approaches for operations and management that cannot keep up with rapid evolution in infrastructure and application demands. InfraAI will develop new, general building blocks and abstractions for AI-driven control of systems writ large, reducing human intervention while ensuring adaptive, high-performance, and safe operations in dynamically changing environments.  

The InfraAI Center is a timely, strategic initiative that builds on UT Austin’s unmatched strengths in systems and AI. With leading expertise spanning networking, OS, and distributed systems, cloud computing, and ML systems, alongside world-class leadership in generative AI, ML theory, and formal methods, UT is uniquely positioned to shape the future of systems infrastructure. 

The InfraAI Center will accelerate and amplify the momentum generated by UT Austin’s premier research institutions, including the Machine Learning Lab (MLL), the high-impact LDOS NSF Expedition in Computing, and the Institute for Foundations of Machine Learning. These efforts have laid the groundwork for transformative advances that InfraAI is uniquely positioned to take to the next level. The newly launched Center for Generative AI and the Texas Advanced Computing Center (TACC) bring world-class capabilities that will supercharge the Center’s research. With strong support from college and university leadership, the InfraAI Center is poised to become a national leader in foundational systems and AI research, driving innovation that shapes the future of intelligent computing.