- A partnership between Oracle and Red Bull Racing has pushed cloud computing into a central role in Formula 1 (F1) race preparation and race-day strategy.
- Cloud computing has allowed Red Bull Racing to run billions of race simulations each weekend.
- Future projects born from race-day analysis include designing a new F1 powertrain for 2026.
At a critical point in the Formula 1 Canadian Grand Prix race in June—with fan-favorite Max Verstappen’s Red Bull Racing car slightly leading Carlos Sainz Jr., who drives for Ferrari—Verstappen’s team had a key decision to make. Just eight laps in, the virtual safety car came out to slow the race cars down, leading to the pivotal question: to pit, or not to pit?
In Formula 1 races, it’s mandatory for drivers to take at least one pit stop. The real question is when to take it. If Red Bull pitted under the virtual safety car, it could reduce time lost during a stop. On the other hand, it would force the team to give up the lead, and keeping Verstappen out on the track could also introduce the risk of deteriorating tires later in the race.
To decide what to do, the team ran simulations. Plenty of them. And in real time.
“If we pit under the virtual [safety] car, we’d give up the lead, but the simulations were confident we would get the lead back again,” Will Courtenay, chief race strategist for Red Bull Racing, tells Popular Mechanics. “We took the pit stop, Sainz stayed out. Now we are 10 or so seconds behind, but then as the simulations correctly predicted, Sainz struggled on tyres, and we got the lead back later in the race.” Verstappen won the Canadian Grand Prix on June 19, continuing his lead in the 2022 F1 standings, the same standings he dominated in 2021. (Teammate Sergio Pérez remains in third). Verstappen’s latest win, the Hungarian Grand Prix on July 31, was also thanks in part to some smart decision-making on which tires to choose.
Courtenay says the power of simulations, due to the new partnership with Oracle, has given the Red Bull team access to more data and better decision-making, both before and during a race. Red Bull went on a search for a cloud-based solution with a global reach, leading to Oracle joining the team in 2021. The partnership went so well that Oracle became a title partner in 2022. That continues to lead to beneficial results, both on the engineering side and in race strategy, according to Zoe Chilton, head of partnerships for Oracle Red Bull Racing (the full name Red Bull is currently competing under).
Every F1 team runs simulations. But Red Bull runs trillions of them per weekend. Under F1’s budget cap, introduced in 2021, moving those simulations off-premise and into the cloud as teams traveled to 22 different countries not only saved Red Bull money it could be spending elsewhere, but sped up the process. “It is not just agility and efficiency,” Chilton tells Popular Mechanics. “By not running on-premise all week, it is a massive savings for us.”
“What we are trying to do is create a simulation of what a real race might look like,” Courtenay explains. “We feed in the model various factors—how tires perform, pace of all the cars, how long it takes to make a pit stop, a good or bad race start—and model as many aspects of the race as we possibly can.”
Then, the team tweaks a variable and runs the simulation again. And again. And again.
“Effectively, we end up doing this over the course of the weekend literally billions of times, generating billions of races and averaging results of all the simulations,” he says. Then, the team experiments with differing strategies. Once the race starts, the simulations don’t stop. The live simulations are where the value add has really come in, according to Courtenay. “Maybe suddenly you have a bad start and dropped six positions and now you’re in a completely different race scenario,” he says. “Our simulations have the capability of constantly updating at every overtake, a finish of a lap, updating predictions and recommendations based on how the race is unfolding. The live simulation is really existing in that situation.”
Take Miami as an example. A new track for the circuit in 2022, uncertainties about tire performance existed well before the race. Once it started, the team quickly saw the tires were holding better than previously thought. With that data fed into the simulation, Courtenay says they were quickly able to pivot to a one-pit strategy.
“You haven’t got the time to sit down and have a good think,” he says. “What is the best thing for me right now? And I need the answer straight away. The data quickly converged to a one-stop [strategy] and that gave us the confidence we could sit it out and go to a one-stop rather than having to panic. In a lot of uncertainty, it really helped us come to a good conclusion as quickly as we could.” On May 8, Verstappen won the Miami Grand Prix, and teammate Sergio Pérez came in fourth.
Moving to Oracle’s cloud-based system allowed Red Bull Racing to increase the number of simulations 25 times while accelerating speed 10 times over. “I get better results when I get them quicker,” Courtenay says. “When the safety car comes out, you need an answer straight away. You can’t wait 30 seconds.”
Come 2026, F1 moves into the next stage of regulations, and Red Bull has embarked on the process of building its own powertrains. Chilton says they’re starting from scratch, so when thinking of engineering engine parts or car design, having the power of continuous simulation is paramount. “It is a very bespoke product for quite the niche,” she says. “If we use Oracle’s cloud infrastructure, we can test what kind of hardware and software run best. We can sample a lot of things. As we go forward, we are hoping to use more cloud resources for more fundamental business processes as well as the strategy simulations.”
The unique demands of F1 have benefits for others. Chilton says pushing the boundaries of what companies can do sets the team up for future success and can help a business explore new strategies.
“Oracle Red Bull Racing is our most demanding customer in terms of short-term demand,” Taylor Newill, senior director of motorsports engineering at Oracle, tells Popular Mechanics. “It has worked out really well and the infrastructure that Will and his team are driving us to build can service customers in financial and healthcare better as well.”
Already, the hardware Oracle used to speed up the race simulations has successfully translated to health sciences customers that require faster data.
Oracle projects don’t stop at simulations and engineering. From small tasks, such as Oracle’s image-recognition system—with machine learning classifying images of rival cars to free up a staff member for more important work—to using AI technology to support the young driver training program, expect the reliance on computing speed to reach every aspect of Red Bull Racing’s business strategy.
Courtenay, though, remains focused on his task: race strategy. He’s already got plans for taking Oracle’s simulations up a notch. “We want to add more complexity, so we get a more accurate model of the race,” Courtenay says. And he needs it all right now. Life in F1 demands it.
This content is created and maintained by a third party, and imported onto this page to help users provide their email addresses. You may be able to find more information about this and similar content at piano.io