Muhammad Abdelghaffar Awad

Muhammad Abdelghaffar Awad is a Principal Member of Technical Staff at AMD Research, where he designs, architects, and leads the development of agentic systems that optimize GPU code and boost engineering productivity, as well as system-level software, libraries, and runtimes for ML and multi-GPU workloads. He designed Iris, an open-source Triton-native multi-GPU programming framework, and tech-leads the multi-team effort building it. He created IntelliKit, an open-source toolkit that gives LLM agents first-class access to GPU profiling and analysis, and IntelliPerf, an agentic performance engineer built on it. He also serves as tooling lead for AMD's company-wide ML for performance engineering initiative and leads cross-organizational efforts spanning teams in ML frameworks, kernel engineering, distributed systems, and research.

He received his M.S. and Ph.D. in Electrical and Computer Engineering from the University of California, Davis, advised by Professor John D. Owens; his dissertation focused on fully concurrent GPU data structures. He received his B.Sc. in Naval Architecture and Marine Engineering from Alexandria University, Egypt, in 2013. His research interests include agentic systems and AI-driven performance engineering, ML systems, GPU and heterogeneous systems, multi-GPU and distributed computing, and kernel programming and optimization. He serves on the review board for the IEEE Transactions on Parallel and Distributed Systems and contributes to the OpenSHMEM specification.