IN THE MIDST OF WORLD WAR I, as modern technology wreaked unprecedented destruction on the battlefields of Europe, a mathematician named Lewis Fry Richardson set out to solve the complex problem of predicting the weather. The seemingly intractable challenge of numerical weather forecasting lay in modeling the earth’s atmosphere with partial differential equations. Richardson estimated that sixty-four thousand “computers” (the term then referred to humans carrying out calculations with slide rules) would be necessary to complete the task. While the prohibitive scale of such daunting operations