Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

HPCwire published news on GPU WRF

stan_posey

New member
Please review an HPCwire article published yesterday on GPU-accelerated WRF:

Meet TempoQuest, a Startup Bringing Weather Code Into the Accelerated Era

NVIDIA supports TQI in GPU development of WRF and the article describes performance results and user benefits.

In particular GPU strong scaling is shown for ORNL Summit where a very large WRF simulation is run up to 512 nodes.
This study originates from a talk at the American Meteorological conference, led by DOE Argonne NL:
Performance Evaluation of the Weather Research and Forecasting (WRF) Model on the DOE Summit Supercomputer
Gökhan Sever, ANL, Argonne, IL; and J. Adie, S. Posey, and C. Catlett

TQI also joined with Microsoft at GTC23 with a talk on their experiences with a wind energy user case and GPU WRF:
Accelerating Wind Energy Forecasts with AceCast [S51589]
Gene Pache
, Founder, President, and CEO, TempoQuest
Amirreza Rastegari, Senior Program Manager, Azure Specialized Compute, Microsoft


Attached are DGX A100 benchmarks that demonstrate GPU WRF with ~30x speedup for a wind energy use case.

Thank you.

--

Stan Posey
NVIDIA
Program Manager
Earth System Science and CFD Solutions Development
sposey@nvidia.com
 

Attachments

  • WRF_TQI_Jan23.pdf
    697.5 KB · Views: 2
Top