Energy of Computing: Unsustainable Trends and Potential Solutions

The current era of computing is driven by mostly general-purpose computing architectures with some specialization. With the increasing use of artificial intelligence/machine learning (ML) applications in manufacturing, natural language processing, scientific applications with increasing requirements such as weather prediction and protein folding, and even in Level 3 & 4 driverless cars, our reliance on computing for enabling these applications is also rapidly increasing. In this presentation, based on examining several trends in computing including energy, complexity of applications, algorithms, and manufacturing, we observe that the current trends are unsustainable. We will try to address the following questions: What can be done to address this increasing appetite for computing which needs increasing amounts of resources to operate? How do we reconcile the much-touted advent of Artificial General Intelligence (AGI) with these current trends? We will conclude the talk with a proposal of a modified form of Turing’s test that points to a new conceptualization of computing for existing and new applications.

Speaker: Sadasivan Shankar, SLAC National Laboratory

Attend in person or online (see weblink)