Scott Beamer

Biographical Sketch

I design architectures for data-intensive applications, with a focus on improving communication efficiency. I am interested in: computer architecture, data movement optimization, memory interconnects, open-source hardware design, and graph algorithms.

I completed my PhD in Computer Science at UC Berkeley while working on the GAP Project. Previously, I was a postdoctoral fellow at Lawrence Berkeley National Lab in the Computer Architecture Group.

Research Interests

My research focuses on improving communication efficiency, since for many data-intensive workloads, accessing the necessary data (communication) is often a bigger challenge than the computation itself. For these communication-bound workloads, optimizing communication can bring many benefits, whether it be improved performance, decreased energy consumption, or reduced manufacturing cost. Much of my work focusses on graph processing, as its data-intensive nature exposes many communication challenges. Going forward, I'm excited to bring my approach to other computational patterns and application domains.

To improve communication efficiency, one can either reduce the amount of communication (move less data) or accelerate the rate of communication (increase utilized bandwidth). My research utilizes both approaches, and I apply them by optimizing software or designing new hardware. To guide these optimization efforts, I analyze the workload to identify bottlenecks. To more aggressively improve hardware communication efficiency, I also investigate the best ways to leverage new technologies such as monolithically integrated silicon photonics.