AI Efficiency
AI workloads are growing in compute and memory demands far faster than hardware capabilities can keep up. We are developing systems techniques and AI-system interfaces that bridge this efficiency gap.
We conduct research in computer architecture and computer systems, with the goal of making computing systems at all scales faster, more energy-efficient, cost-effective, and secure. Our recent work focuses on cloud computing infrastructure, systems for machine learning, and the use of machine learning to optimize systems.
AI workloads are growing in compute and memory demands far faster than hardware capabilities can keep up. We are developing systems techniques and AI-system interfaces that bridge this efficiency gap.
Today’s AI infrastructure, rooted in HPC paradigms, is inflexible and expensive. We are designing AI systems around scale-out principles to improve scalability, resilience, and cost effectiveness.
We are co-designing hardware, system software, and applications to improve the efficiency of cloud infrastructure, frequently using AI techniques to automate optimization.
2025-05-28
Congratulations to Dr Mark Zhao on submitting his PhD thesis titled “Scalable, efficient, and secure machine learning data systems”. Mark is joining the faculty at Boulder CS in the fall.
2025-05-27
Congratulations to Dr Timothy Chong on submitting his PhD thesis titled “Addressing endpoint-induced congestion with duplicate acknowledgment”.
2025-05-27
Congratulations to Zhiqiang Xie on being awarded an NVIDIA Graduate Fellowship.
Grant Ayers
Woongki Baek
Adam Belay
Timothy Chong
JaeWoong Chung
Michael Dalton
Christina Delimitrou
Mingyu Gao
Samuel Grossman
Rehan Hameed
Jack Humphries
Kostis Kaffes
Hari Kannan
Ana Klimovic
Jacob Leverich
Qian Li
David Lo
Austen McDonald
Chi Cao Minh
Camilo Moreno
Raghu Prabhakar
Suzanne Rivoire
Francisco Romero
Daniel Sanchez
Geet Sethi
Yawen Wang
Sewook Wee
Richard Yoo
Mark Zhao
Ahmad Zmily
Dimitris Economou
Heiner Litz
Daniel Mendoza
Travis Skare
Neeraja Yadwadkar