[cepth]

In 2017, Ars Technica did a deep dive into computation in Formula 1 (https://arstechnica.com/cars/2017/04/formula-1-technology/).

Some relevant quotes:

> For example, each Formula 1 team is only allowed to use 25 teraflops (trillions of floating point operations per second) of double precision (64-bit) computing power for simulating car aerodynamics.

> Oddly, the F1 regulations also stipulate that only CPUs can be used, not GPUs, and that teams must explicitly prove whether they're using AVX instructions or not. Without AVX, the FIA rates a single Sandy Bridge or Ivy Bridge CPU core at 4 flops; with AVX, each core is rated at 8 flops. Every team has to submit the exact specifications of their compute cluster to the FIA at the start of the season, and then a logfile after every eight weeks of ongoing testing.

> Everest says that every team has its own on-premises hardware setup and that no one has yet moved to the cloud. There's no technical reason why the cloud can't be used for car aerodynamics simulations—and F1 teams are investigating such a possibility—but the aforementioned stringent CPU stipulations currently make it impossible. The result is that most F1 teams use a somewhat hybridised setup, with a local Linux cluster outputting aerodynamics data that informs the manufacturing of physical components, the details of which are kept in the cloud.

> Wind tunnel usage is similarly restricted: F1 teams are only allowed 25 hours of "wind on" time per week to test new chassis designs. 10 years ago, in 2007, it was very different, says Everest: "There was no restriction on teraflops, no restriction on wind tunnel hours," continues Everest. "We had three shifts running the wind tunnel 24/7. It got to the point where a lot of teams were talking about building a second wind tunnel; Williams built a second tunnel.

With the new cost cap in F1 (https://www.autoweek.com/racing/formula-1/a35293542/f1-budge...) (which notably excludes driver salaries), it would be interesting to know how much these on-prem clusters cost to operate.

[Someone]

To level the field even more, I think the FIA should require teams to release the design of their computer hardware after X time. That way, investments by one team on improving the system architecture spread to teams with lower budgets after a while.

Also, I didn’t find it in the article, but I guess they have programmers who can work for months to speed up their software by a few percent.

[RaceWon]

>> to level the field even more, I think the FIA should require teams to release the design of their computer hardware after X time.

But that is Not what Formula One is about... it is Not a Spec series where the cars are equal to each other. It is a competition where each team builds their own race car to compete against the other iterations of race cars built by the opposing teams. It is Not meant to be fair or equitable. We have Indycar and NASCAR for that.

Ditto with the drivers: Is Max or Lewis comparable to say a Mazepin or even a Hulkenberg? No they are Not.

It's a Spectacle, it's a Circus... that is what F1 is about. And I tell you, as a racer there is nothing else that is its equal in terms of pure audacity both from a standpoint of driving talent and car performance.

[moralestapia]

>each Formula 1 team is only allowed to use 25 teraflops

That ... doesn't make much sense, honestly.

[jecel]

The problem is the sloppy use of technical terms. MIPS means "millions of instructions per second" so the "p" is "per" and the "s" is "second". So it is natural for people to use FLOPS in the exact same way, but it is more correct for this to be "FLoating point OPerationS" where the "s" is used to indicate a plural.

That makes MIPS the equivalent of power (Watts) and FLOPS the equivalent of energy (Jouls). In that case limiting each Formula 1 team to a maximum of 25 teraflops of computation does make sense.

If instead you use teraflops as the equivalent of "trillion floating point operations per second" as many people do then it indeed makes less sense.