Formula One and AWS Use HPC to Build a Better Racing Car

New design will allow more efficient drafting in wheel-to-wheel racing scenarios

Formula One Group (the company behind the Formula One racing car) and Amazon Web Services have completed a computational fluid dynamics study to design the Formula One racing car for the 2021 racing season. The study required over 12,000 hours of compute time and represents the first in-depth aerodynamics study that accounts for how cars interact with each other when racing.

According to the press release, the aerodynamics of a Formula 1 car generates a downward force that allows the car to hug the road and travel faster through curves. However, a car that is following close behind another car (within one car length) loses 50% of that downward force. The study looked for ways to reduce the loss in downward force with wheel-to-wheel racing. The project used 1,150 cores and based the simulations on 550 million data points. Insights discovered in the simulations led to a new design that reduces the lost in downward force to 15%.

“This project with AWS was one of the most revolutionary in the history of Formula 1 aerodynamics,” said Pat Symonds, Chief Technical Officer of Formula 1. “Nobody designs a car to come in second, but for this CFD project we were looking at how cars perform in the wake of another, as opposed to running in clean air. We have been able to use AWS technologies to understand the incredible aerodynamic complexities associated with multi-car simulations, and are pleased that the results indicate we have made excellent progress towards our aims of closer racing.”