Nathan Beckmann

Nathan Beckmann

Associate Professor

Website

ORCiD

Office 9021 Gates and Hillman Centers

Email beckmann@cmu.edu

Phone (412) 268-7412

Department
Computer Science Department

Administrative Support Person
Michael Stanley

Biography

Nathan Beckmann is an Associate Professor in the School of Computer Science at Carnegie Mellon University. He received his PhD from MIT in 2015 advised by Daniel Sanchez, his MS from MIT in 2010 advised by Anant Agarwal, and BS degrees in 2008 from UCLA in EECS and Mathematics. His PhD dissertation was recognized with the George M Sprowls Award for an outstanding dissertation in electrical engineering and computer science at MIT. He has received multiple Best Paper awards, the NSF CAREER Award, a Google Research Scholar Award, and a Sloan Research Fellowship.

Research/Teaching Statement

My research builds energy-efficient, general-purpose computers. General-purpose computing is the most impactful technology of the last fifty years, and it is essential that its growth continues. Unfortunately, general-purpose computing is currently under threat. The future of general-purpose computing depends on energy efficiency — from IoT to datacenters, energy efficiency determines computers’ capability, lifetime, and environmental impact. General-purpose computer architectures have stagnated, and their energy efficiency has not significantly improved for over a decade.

Consequently, the last decade has seen a dramatic shift towards specialized hardware. Hardware specialization has delivered large gains in energy efficiency, but at a loss of general-purpose programmability and dramatic increase in cost and carbon footprint. The challenge for the next decade is to reconcile programmability and energy efficiency to enable a truly sustainable future for computing.

My research improves the energy efficiency of general-purpose computers by orders of magnitude. My research group is designing spatial dataflow architectures that avoid the inherent inefficiency of today's von Neumann CPUs while keeping computation near-data to mitigate the growing cost of data movement. This research spans hardware and software, with a particular focus on compilers and computer architecture. My focuses on build complete systems, including code that has run in production data centers and three silicon testchips.