PDA

View Full Version : Tokyo Tech Weather Forecasting Model Gets 80X Perf Boost Through GPU Acceleration


News
03-30-10, 02:00 PM
Quick and accurate predictions of climate and weather are a part of everyday planning - think staying dry while walking to work - all the way¬*to the serious, large-scale initiatives - like knowing what¬*resources to allocate¬*for hurricane recovery efforts.

http://blogs.nvidia.com/.a/6a00d834515fca69e201310ffbdbbe970c-320wi (http://blogs.nvidia.com/.a/6a00d834515fca69e201310ffbdbbe970c-pi)

GPU technology has already had a hand in improving the precision of weather forecasting. In Boulder, Colorado, the National Center for Atmospheric Research (http://www.nvidia.com/object/national_center_for_atmospheric_research.html) contributed to the development of the Weather Research and Forecasting Model (WRF) which partially utilizes GPU technology for¬*up to¬*20% speed improvement.¬* WRF is the is the most widely-used model in the world,¬*counting the¬*National Weather Service and the Air Force Weather Agency among its users.

Now, we've reached another¬*milestone for the GPU in the field of atmospheric modeling. A¬*research group led by Professor Takayuki Aoki of theTokyo Institute of Technology (http://www.titech.ac.jp/english/) has succeeded in 100% utilization of GPUs in the next-generation weather forecasting model, codenamed ASUCA, currently being developed by the Japan Meteorological Agency (http://www.jma.go.jp/jma/indexe.html). ASUCA has a similar feature set to WRF, but because it is fully GPU-optimized, ASUCA runs 80 times faster than weather models running on CPUs alone or on CPU/GPU combinations. In short, it is the fastest solution available today.



Thanks to NVIDIA Tesla (http://www.nvidia.com/object/tesla_computing_solutions.html) and CUDA parallel processing architecture (http://www.nvidia.com/object/cuda_home_new.html), ASUCA simulates a 6 hour event (with 2km mesh size in a 3164x3028x48 grid) in 70 minutes on 120 GPUs, a calculation that would have taken 5600 minutes using CPUs. The speed and accuracy of these weather predictions have major implications for more efficient and effective planning and resource allocation.

http://blogs.nvidia.com/.a/6a00d834515fca69e20133ec55d61d970b-320wi (http://blogs.nvidia.com/.a/6a00d834515fca69e20133ec55d61d970b-pi)

It‚??s an exciting time for the GPU in atmospheric science, and we want to congratulate Professor Takayuki Aoki and his team on their hard work and incredible breakthrough.

http://feeds.feedburner.com/~ff/nTersect?d=yIl2AUoC8zA (http://feeds.feedburner.com/~ff/nTersect?a=XyeSIIgDYlI:VokqrJET3eY:yIl2AUoC8zA) http://feeds.feedburner.com/~ff/nTersect?i=XyeSIIgDYlI:VokqrJET3eY:V_sGLiPBpWU (http://feeds.feedburner.com/~ff/nTersect?a=XyeSIIgDYlI:VokqrJET3eY:V_sGLiPBpWU)
http://feeds.feedburner.com/~r/nTersect/~4/XyeSIIgDYlI

More... (http://feedproxy.google.com/~r/nTersect/~3/XyeSIIgDYlI/tokyo-tech-weather-forecasting-model-gets-80x-perf-boost-through-gpu-acceleration.html)