The United States, Europe and China have all embarked on an ambitious program to build an exascale computer (10^18 ops/second). These computers will rely upon massive (and unparalleled) levels of parallelism in attempt to stave off the inevitable death of Moore's law. This program has an unremitting focus on hardware, and software that gives practical access to these machines. As the attention draws away from the hardware, the technical work also recedes.  While the use of parallelism and ancillary technical issues provide some challenges for the mathematics community, the span and scope of the challenge is rather restricted compared to historical contributions of mathematics. Taking a look back at the history of mathematics contributions to high performance computing may be useful in crafting a more balanced and ultimately successful view. High performance computing owes a great deal the mathematic's community and the contributions to the current program should be much greater than currently envisioned. Fluid mechanics offers many timeless and inspiring examples that help inform this more balanced and visionary path for computing beyond simply chasing "big iron".

Refreshments will be served in the lounge at 3:00.