Divya Mahajan gave an invited talk at the 2026 Grainger Foundation Frontiers of Engineering Symposium (US FOE), held at the University of Texas at Austin.

This year’s symposium brought together 100 early-career engineers, with Mahajan among just 14 participants invited to give a talk.

Mahajan, a Sutterfield Family Early Career Professor in the School of Electrical and Computer Engineering with a joint appointment in the School of Computer Science, spoke as part of the Compute Challenges for Artificial Intelligence (AI) session. Her talk, “Scaling Intelligence Beyond the Infrastructure Wall: Cross-Layer Co-Design for Efficient and Adaptive AI Infrastructure.

First, she discussed how cross-layer co-design can push the efficiency frontier, jointly rethinking AI workloads, systems, and hardware rather than optimizing each layer independently.

Second, she showed how adaptive systems can maintain that frontier, responding to changing hardware characteristics, workloads, and operating conditions to sustain efficient execution over time. These ideas are also central to her recent NSF CAREER Award.

“One of the most exciting parts of Frontiers of Engineering was seeing researchers from so many different disciplines thinking about AI and its challenges,” Mahajan said. “AI now touches almost every field, but each community approaches it from a different perspective. Bringing those perspectives together led to conversations and connections that would be difficult to have within any one discipline.”

Organized by the National Academy of Engineering, US FOE brings together highly accomplished early-career engineers from industry, academia, and government for a three-day, invitation-only symposium. The meeting fosters interdisciplinary exchange around emerging engineering challenges and opportunities for collaboration among the next generation of engineering leaders.

Mahajan leads the Systems for Scalable Intelligence (SCAI) Lab at Georgia Tech. Her research focuses on building scalable and sustainable computing systems for AI, spanning computer architecture, systems, and data management. Her group develops cross-layer approaches to push the capabilities of AI infrastructure while making systems adaptive to the hardware and operating environments on which they run.