Optical Interconnects and Extreme Computing

Keren Bergman, John Shalf and Tom Hausken

The leap to “exaflop” computing will require new optical technologies—and new co-design approaches to developing them.

figure

[iStock]

As they confront ever more complex and data-intensive problems, scientists and researchers increasingly look to the next generation of supercomputing—the high-end segment of high-performance computing (HPC). That next generation will play out in so-called exaflop computers—machines capable of executing at least a quintillion (1018) floating-point operations per second (flops). Such a computer would represent a thousand-fold improvement over the current standard, the petaflop machines that first came on line in 2008. But while exaflop computers already appear on funders’ technology roadmaps, making the exaflop leap on the short timescales of those roadmaps constitutes a formidable challenge.

Log in or become a member to view the full text of this article.


This article may be available for purchase via the search at Optica Publishing Group.
Optica Members get the full text of Optics & Photonics News, plus a variety of other member benefits.

Add a Comment