 |
Center for Nonlinear Studies |
Why do computers crash, often at the slightest provocation? Living organisms often suffer major systemic insults yet continue to run. Could we build a machine that didn't crash -- even one that -- couldn't -- crash? In this talk -- mixing revisionist history and overt advocacy with extreme computer architecture and new programming models -- I will argue that we in computing have been optimizing the wrong thing, and suggest that the route to truly large-scale computing places robustness first.
Host: Garrett Kenyon, gkenyon@lanl.gov, 7-1900, IS & T