Oracle Red Bull Racing will start the 2026 Formula 1 season with an engine it developed itself. It has partnered with Ford to do this. The result of this collaboration is Red Bull Ford Powertrains, which is developing this completely new hybrid power source for Oracle Red Bull Racing’s Formula 1 cars. How big is this undertaking for Oracle Red Bull Racing? And what role does the IT environment play in this development? We put these and other questions to Matt Cadieux, Oracle Red Bull Racing’s CIO.
Building an internal combustion engine is not the easiest thing to do. However, if that engine is intended for use in Formula 1 Grands Prix, it gets even harder. Starting this is madness if you don’t already have a lot of experience building an engine. Yet that is exactly what Oracle Red Bull Racing decided to do several years ago. Red Bull Ford Powertrains was thus born and the development of the engine with which Oracle Red Bull Racing (and Visa Cash App RB) will race from season 2026 began.
Why is Red Bull Ford Powertrains building its own engine?
Oracle Red Bull Racing has been active in Formula 1 for many years under different names. Until 2022, however, the team did not develop its own power unit. It procured those, most recently from Honda, but also in the past from other manufacturers. The team went to work with the purchased engine to squeeze the most out of it. However, in August 2022, the organization behind Formula 1 made a big decision. Starting in 2026, Formula 1 cars need to have hybrid engines whose Co2 emissions were compensated. This included integrating a more powerful electric motor and the use of renewable fuels.
The new power unit that Formula One teams must have in their cars by 2026 had to be built, tested and validated in about four years. Despite the decrease in complexity of the new engine by omitting certain parts that are still in the current ones, that was still quite a task. That was and is true for established engine manufacturers, and thus all the more so for Red Bull Ford Powertrains.
Cadieux therefore calls Oracle Red Bull Racing’s choice (in the person of Christian Horner) to develop the engine itself “a very brave decision.” After all, there was in fact no knowledge (“zero knowledge,” in Cadieux’s words) inside Oracle Red Bull Racing about how to go about doing this. So it was not easy to say beforehand with any certainty that the team would succeed in doing this.
Brave decision, but was it also a smart decision?
To say that it was a brave decision to develop an engine themselves actually seems like an understatement to us. Sure, Oracle Red Bull Racing had a lot of experience and expertise in developing a chassis for Formula One cars, but that is not nearly as complex as developing an engine.
“From an IT perspective, there was a lot of experience with simulations for building a chassis, but there was no idea what this was like in developing an engine. The idea was that there should undoubtedly be overlap,” Cadieux outlines the mindset among Oracle Red Bull Racing people at the beginning. In practice, however, that turned out to be slightly different, he also readily admits. “Powertrains are a much bigger animal,” according to him. In particular, it involves Computational Fluid Dynamics (CFD). That’s one of the main parts to get right when developing a new engine. If you get that right, you know exactly how an engine performs, or at least you can estimate it very well. In Cadieux’s words,however, this results in a “large amount of complexity.”
In itself, a large amount of complexity is not necessarily a problem. The problem at Oracle Red Bull Racing was mainly that there were no business processes, no IP and no infrastructure to properly get started with CFD. They still had to set up a team to work with it anyway. So that was the first step. Two people were hired who knew what it took to get started.
The Oracle in Oracle Red Bull Racing
Pretty soon after the project to develop a proprietary power unit really took off, Oracle came into the picture. More specifically, Oracle Cloud Infrastructure (OCI) caught the eye of Oracle Red Bull Racing. That is, the team needed High Performance Computing (HPC) when developing an engine. OCI can deliver that. On the other hand, however, the team had no idea how much HPC it needed. That is why the cloud made sense to them. “In the cloud, we could start small. As we developed our tool stack, we could also experiment more,” Cadieux said. That’s important for the initial investment, because now he and his team didn’t have to estimate the number of cores needed, but could scale up as needed.
Of course, Oracle Red Bull Racing had and has a lot of on-prem infrastructure. That was and is suitable for specific things (more on that later), but cannot run HPC workloads at scale. “We also did not have enough power available, even if we could have done it,” Cadieux adds. It would have taken at least a year to upgrade the connections. “That is, if we had been able to estimate at all what we would have needed,” he underlines again how unclear and uncertain everything was at the beginning of the process.
Thanks to OCI, Oracle Red Bull Racing and Red Bull Ford Powertrains can compete with other teams when it comes to building a hybrid power unit for the 2026 season. “Without Oracle, we would have had to extend the whole process by at least a year,” Cadieux points out. Now the team could give the developers/engineers something to work with within a few weeks. It also helped that Oracle already had a relationship with the main solver used by Oracle Red Bull Racing for CFD. That is the one from Converging Science, a company focused on CFD. One and one is more than two in this case.
Hybrid model
OCI made it possible for Oracle Red Bull Racing to be up and running quickly. However, simply saying “just give us a few pounds of OCI” is obviously not enough. It also needed the people to actually carry out all the work. “In the first 12 months, the team went from two to 10 people,” Cadieux outlines the progress. During that time, there was a lot of tinkering with configurations and the settings for the CFD solver the team used. Consequently, the models deployed after 12 months were radically different from those at the beginning.
Initially, Oracle Red Bull Racing was on the familiar “pay as you go” model that the cloud is known for. However, that also provided uncertainty. So over time, the team gave Oracle a growth plan. This is important because it can then be assured of sufficient local capacity, in London. Cadieux even calls this “crucial” because the company has a hybrid model. That is, compute is in OCI, as well as storage and memory. However, the results of the simulations done on the HPC clusters in OCI go to on-prem locations. There, Oracle Red Bull Racing has a large parallel file system. This link must be as fast as possible. They have a 10GE private line from AT&T to the team’s data center for that.
There are two reasons why Oracle Red Bull Racing has not transferred everything to OCI (yet) and is still keeping some of the storage on-prem. On the one hand, there is a practical consideration. The software it uses for post-processing is not developed for use in the cloud and they didn’t have time to transfer it. Also, it’s just nice to have (and feel) control over the data and to know that the post-processing is working as it should. Cadieux calls this the “sweet spot” during our conversation.
Full focus on complexity, not HPC
For Oracle Red Bull Racing, OCI primarily provides the focus necessary to solve CFD complexity. “It’s not just about the airflow, but also the explosions of the fuel and so on,” Cadieux points out. That means he doesn’t want to use his people to manage HPC clusters, but to solve these kinds of complex issues. You can’t do that if you want to run everything on-prem.
In addition to the technical challenges that F1 teams face, there are increasingly financial challenges these days. In fact, a cost cap has been introduced. For the chassis that was the case in 2021, for the power unit in 2023. Obviously, the people at Oracle Red Bull Racing have been well informed about this.
The cost cap means, among other things, that spending on IT must be kept to a minimum. “F1 teams focus on squeezing as much as possible out of a car, anything they can save on IT is a bonus,” in Cadieux’s words. Sometimes that can also mean that CAPEX in OCI yields more than OPEX. Or in other words, sometimes it’s good to buy certain things upfront, including in the cloud, and not go for a pay-per-use model. In the end, Cadieux is very satisfied about the financial side of things: “OCI delivers great value for money.”
The big question: will Red Bull Ford Powertrains have a good engine in 2026?
The development of a new power unit has undoubtedly been a learning process for Oracle Red Bull Racing (and Red Bull Ford Powertrains). That is only logical, considering the fact that it didn’t have any experience with building an engine. The most important thing for the team is to have a competitive engine in the Red Bull cars in 2026. When asked if this will be achieved, Cadieux is optimistic. “We are meeting our targets in terms of output and reliability,” he indicates.
In addition to the targets Oracle Red Bull Racing hits, he also feels that the challenges his team face not much different from those of other teams, even though they may have (much) more experience in building engines. They have never done this before either, is his argument. And Oracle Red Bull Racing, of course, also hired a number of specialists to build the new engine and entered into a technical partnership with Ford. They did have prior experience.
All in all, Oracle Red Bull Racing is on track, according to Cadieux, to be able to put an engine in the Red Bull cars by 2026 that can compete for the Formula 1 crown. If they succeed in doing this, that is a big achievement. An achievement that probably would not have been possible without the cloud, in this case the Oracle Cloud. Every engine that Oracle Red Bull Racing will put in a Formula One car starting in 2026 will be a hybrid engine running on fuel, electricity but also a sprinkling of cloud.
Also read: Can the right cybersecurity solution make Verstappen’s Red Bull go faster?