Eugene P. Pache Jr., PresidentSeasonal changes, as well as daily and weekly fluctuations in weather can create unprecedented losses to firms across industries such as Retail, Aviation, Utilities, and Transportation among others. Mesoscale Forecast, which applies Numerical Weather Prediction to limited areas using high resolution, is gaining prominence in this situation. To enhance accuracy of Mesocale Forecasts, TempoQuest, a Charlotte, NC, based SaaS company, is developing weather forecasting software—utilizing NVIDIA’s GPUs which reduces the data processing time by a factor of ten. “If the data processing speed can be accelerated, then more datasets from different sensors can be used which will then improve the accuracy of the forecast,” says Eugene P. Pache, Jr., President, TempoQuest. “With the rapid processing, forecast can be disseminated at a much higher frequency.”
The firm’s suite of software solutions are being designed to run on GPUs instead of CPUs, dramatically increasing the speed and accuracy of analysis— technology that has not yet been applied to comprehensive weather data analysis. “We are developing a suite of software products that will leverage the power of NVIDIA’s GPU accelerators and CUDA coding, which will enable our clients to make better business decisions,” says Pache. To deliver precise forecasts, TempoQuest has collaborated with NVIDIA Corporation for their technical support on software integration, application engineering, and distribution assistance for the firm’s GPU forecasting product.
TempoQuest is still in the R&D stage and the firm’s product is thirty percent complete. The firm leverages the Weather Research and Forecasting (WRF) model, which is an advanced Mesoscale numerical weather prediction system. It features multiple dynamical cores, a 3-dimensional variational (3DVAR) data assimilation system, and a software architecture allowing for computational parallelism and system extensibility. “The WRF model is written in FORTRAN and ported to Computer Unified Device Architecture (CUDA); the extensive CUDA coding maximizes the output of GPUs,” informs Pache. This whole process dramatically increases the speed and accuracy of weather analysis. “TempoQuest is developing a series of products based on the WRF model which will enable data casting by processing and assimilating data into the model,” says Pache.
We are developing a suite of software products that will leverage the power of NVIDIA’s GPU accelerators and CUDA coding, which will enable our clients to make better business decisions
Besides being the pioneer in utilizing GPU accelerators in weather forecasting domain, the extensive knowledge in atmospheric science, weather forecasting models, and high performance computing has allowed the firm to add value to their client’s business.
For commodity traders, insurance firms, major agricultural companies, among others, speed and accuracy are critical to the success of their businesses and an extremely valuable asset. Pache informs that the firm’s forecasting solution can play a big role inthese industries by providing accurate weather forecasts. For instance, a utility company can leverage the Mesoscale weather forecasting solution to optimize their usage of renewable sources. The application can also be leveraged by the aviation industry to optimize their fuel usage and save the cost of re-scheduling and re-routing flights
As currently practiced, TempoQuest will continue to process the data sought from drones, satellites among others, and provide insights to weather companies. TempoQuest’s GPU accelerator and CUDA coded Mesoscale weather forecast solution will open up huge opportunities for companies to improve their performance and profitability. The company also plans to develop a product that will improve the accuracy of forecasting hurricanes. “We will also focus on improving the analysis of air quality in Indonesia and China which will add to the different areas and products we are capable of delivering,” concludes Pache.