Today TempoQuest announced that it has achieved a breakthrough 700% acceleration of the weather community’s Weather Research Forecast (WRF) model used extensively for real-time forecasting and weather research throughout the world.
The acceleration of regional weather processing times has presented a challenge to forecasters that has taken years to solve,” said TempoQuest President Gene Pache. “Current analytical processes are decades old and until the availability of recent computer acceleration technologies pioneered by NVIDIA, and software platforms like AceCAST, there simply hasn’t been the ability to analyze all the data that drives forecasting speed and accuracy.”
The game-changing speeds are the result of TempoQuest’s technology leadership in GPU development of weather modeling software combined with NVIDIA Tesla V100 GPUs to produce TempoQuest’s customizable and proprietary AceCAST software platform for GPU computing.
Why is this milestone important?
- A forecast requiring 60 minutes to produce by best available current forecasting technologies can now be completed in less than 10 minutes
- The implications for saving lives, protecting property and ensuring the profitability of weather-related businesses (e.g., agriculture, renewable energy, automated aerial vehicles and ground transportation) are profound.
High-quality weather forecasts require a tremendous amount of computing speed,” said Stan Posey, Program Manager for Earth System Science at NVIDIA . “We look forward to continued collaboration with TempoQuest that will bring further breakthroughs to the weather modeling community.”
