A post on NVIDIA’s blog site this week spotlights a rather unique application of the company’s GPU hardware and computing technology. According to NVIDIA, CUDA GPUs are being tapped to optimize flight planning for air traffic controllers, the idea being to make airline travel safer, and with fewer delays and congestion.
At any given moment, there are around 7,000 aircraft flying in the skies over the United States, a number which is expected to double by 2025. Tracking those aircraft with human controllers, even with the help of current traffic control software, is a laborious, not to mention error-prone, process. If you’ve spent anytime in an airport lately, you’ve probably experienced some of the inefficiencies of flight planning firsthand.
The baseline air traffic control software, developed by NASA, takes about 10 minutes to perform a 24-hour trajectory prediction in four dimensions (latitude, longitude, altitude and time). That’s assuming 35,000 aircraft will be flying that day. With the CUDA prototype version, the application took less than 2.5 seconds — a speedup of 250X.
The CUDA port of the trajectory software is being developed by NASA and Optimal Synthesis, a Silicon Valley research firm that specializes in various types of software for engineers and scientists. According to Monish Tandale, a research scientist at the company, being able to compute trajectories with CUDA technology has a number of advantages. In an interview published by NVIDIA, he described it thusly:
GPU computing allows us to exploit the parallelism in the trajectory prediction process to have extremely fast run-times. This in turn allows us to achieve real-time performance and analyze models with greater complexity, and opens up the possibility of utilizing algorithms and approaches that were earlier deemed impractical due to the computational complexity.
Monish maintains that porting the code to GPUs was relatively easy thanks to the shallow learning curve of CUDA, allowing them to “smoothly transition from C/C++ based CPU programming to GPU programming.” His company is currently looking to exploit GPUs for other government research projects with NASA, the Missile Defense Agency and the Air Force.