Office 9109 Gates and Hillman Centers
Phone (412) 268-3064
Computer Science Department
Administrative Support Person
David Andersen is a professor in the Computer Science department at Carnegie Mellon University. He received his Ph.D. and M.S. degrees from MIT, and and received B.S. degrees in Computer Science and Biology from the University of Utah. He is founder and CTO of BrdgAI.
He previously served as a visiting scientist in the Google Brain group; co-founder of deep learning startup Marianas Labs; co-founder and CTO of an Internet Service Provider in Salt Lake City; a biology lab assistant; and a rock climbing instructor.
My research focuses on networks and distributed systems. My interests lie in creating systems that meet goals such as robustness, high availability, and energy efficiency. My current research encompasses two large project areas:
FAWN: Fast Arrays of Wimpy Nodes. The FAWN project aims to develop computational cluster architectures, and software techniques to use them, that are drastically more energy and cost-effective than today's technologies. We do so by building clusters from systems that are comparatively slow by the standards of today's leading-edge, but that together, can provide drastically more throughput and computational capability at lower power. We then tackle the problems of actually using such systems by developing new algorithms and systems techniques for data-intensive processing on these clusters. Typical challenges we seek to overcome are using substantially less memory than prior approaches, harnessing new storage technologies such as Flash and phase-change memory, and coping with systems composed of 5x more nodes than previously used.
XIA: The Expressive Internet Architecture. XIA is a clean-slate approach to Internetworking, in a collaboration with numerous faculty at Carnegie Mellon and elsewhere. The goal of the project is to develop an Internet architecture for the next 100 years: one that is robust, has drastically improved security compared to todays, and, most importantly, one that can evolve easily to meet as-yet-unknown uses and challenges.