At the International Supercomputing Conference (ISC) earlier this week, Intel committed to delivering a complete technological solution for ExaScale computing performance later in the decade.
The world's fastest supercomputer is the K Computer in Japan, built by Fujitsu, delivering over 8 PetaFLOP/s performance. It consumes almost 10 Megawatts of power and costs about $10 million a year ... [ read the full article ]
Please read the original article before posting your comments.
So in theory, this should mean that since the "Exascale" computing will improve by 100 times (or a 100%) then WEATHER FORECASTING will become 100% more accurate??????? Because it just sucks today and is only 20% accurate.
Originally posted by hearme0: So in theory, this should mean that since the "Exascale" computing will improve by 100 times (or a 100%) then WEATHER FORECASTING will become 100% more accurate??????? Because it just sucks today and is only 20% accurate.
No. At least, i think so. Weather predictions are still programed by humans(and you never will be able to predict the weather 100%). The weather programs are just simulations. The weather man/woman, input data,(or the radar inputs data, thru the scan). Then the program just "predicts" or "simulates" the weather. Therefor I dont think the more "power" processor or even a better programed Program will ever be able to predict the weather 100%.
Originally posted by hearme0: So in theory, this should mean that since the "Exascale" computing will improve by 100 times (or a 100%) then WEATHER FORECASTING will become 100% more accurate??????? Because it just sucks today and is only 20% accurate.
No not exactly.
Firstly, on the point about improving "by 100 times", what they mean is that an ExaScale machine would be 100 times faster in raw computing power than today's fastest (K Computer in Japan at 8 PFLOP/s, which will reach 10PFLOP/s), that's just going by the FLOP/s alone and there are many other things to take into account such as advancements in programming and simulation models.
The point about weather prediction is not really to tell you when it's going to rain and when it's not with more accuracy. Instead, supercomputing can help to make much better predictions for big events, such as the formation of a hurricane and where it will make landfall.
NASA goes further than this and things that with the right data input we could actually have incredibly accurate weather and climate information, but they are talking on a scale 100 greater than even ExaScale (ZettaScale) so that's a bit longer off.
Originally posted by hearme0: So in theory, this should mean that since the "Exascale" computing will improve by 100 times (or a 100%) then WEATHER FORECASTING will become 100% more accurate??????? Because it just sucks today and is only 20% accurate.
No. At least, i think so. Weather predictions are still programed by humans(and you never will be able to predict the weather 100%). The weather programs are just simulations. The weather man/woman, input data,(or the radar inputs data, thru the scan). Then the program just "predicts" or "simulates" the weather. Therefor I dont think the more "power" processor or even a better programed Program will ever be able to predict the weather 100%.
Actually.........you're right about humans inputting but wrong about more accuracy. More computer power will take more data regarding ALL THE LITTLE TINY WEATHER FACTORS and evaluate. Right now, there is too much data for simple computers to accurately gauge and Dopler is too generalized. Besides, Dopler isn't based solely off human inputting. It's bouncing signals and calculating itself.
