4 ms·
Back in the seventies and eighties there were four sources of impetus for improvement in highly parallel supercomputing. Military/Nuclear, high energy physics,
by ggm 8y ago
Back in the seventies and eighties there were four sources of impetus for improvement in highly parallel supercomputing. Military/Nuclear, high energy physics, oil and gas exploration (seismic analysis) and weather.
The weather people were the nicest. They could talk to almost anyone about their work!
I remember somebody in csiro talking to me about models getting accurate at the 1km square granular scale for a day. I think we're well beyond that for both cell size and duration now.
Everyone loves to talk about the weather and Bob Dylan was wrong about not needing a weatherman to know which way the wind blows