Congratulations to Jonas Axboe for winning and completely outclassing his rivals in the North America Radical Cup! For those who don’t immediately recognize the name, father Jens Axboe is a well known Linux kernel performance engineer, in particular known for the io_uring API. But what is even cooler, Jonas was sponsored by database startup TigerBeetle. If the name sounds familiar, it’s because TigerBeetle’s continuous benchmarking performance data is tracked live at Nyrkiö.com. (Their dashboard can be boring at times as they rarely have any performance regressions!)

Car racing and software performance engineering have a lot in common, making the two a great combination for sponsorship and branding collaborations. In both cases you employ technology to go fast, and obsess about milliseconds. But it goes deeper than that. What they teach young drivers on the racing track is like straight out of a performance engineering text book*:

• Tuning your car is an iterative process of: configure, test, measure, analyze, (repeat).
• You can only change one thing at a time. Because if you change two things, and you get a good lap time, you won’t know which of the change it was that made a difference.
• The paddock is full of well meaning people that will give you helpful advice on how to make your car go faster. DO NOT LISTEN to them. The only way to make your car go faster is to configure, test… (see point 1)
• The clock is the ultimate truth. Sometimes your driving may feel fast, but once you finish the lap, you see the time is actually worse. Your feelings don’t matter. The clock is the truth.
• From which is derived the saying “the clock likes it”

* It’s a figure of speech. There aren’t many text books. We learn our craft by doing and from each other.

As a reminder, the word “benchmarking” originates with the practice of attaching a car engine to a workbench in order to measure its power and torque.

We owe a lot to car racing. If you stay out of the Formula 1 – 4 business, in karting and various so called “standard” car* series, such as GT, TCR or V1600, drivers are often happy to take sponsor packages for as little as 500 – 1000 €. If you want to paint a car in your brand colors like TigerBeetle did, we are talking about tens of thousands.

* There’s nothing standard about them!

Racing is also one of the few sports where men and women compete together in the same classes, on equal terms, making it a very modern sport to be involved in. If you’re interested in sponsoring your own talented and fast future champion, I know many great drivers in Finland and Europe.

At Nyrkiö we have sponsored 3 drivers during our first years 2024 and 2025. The first driver we sponsored, Meri Levula won the Finnish track sprint junior championship this Summer! (This might be the first female champion in Finnish national racing series? If so, truly historic and we are proud to have supported you on the way.) In 2023 Meri also participated in the final tryouts for the FIA Rising Stars program, from which one driver now races in the Formula Academy series, of which there’s also a Netflix documentary series.

On the karting track Nyrkiö sponsored in 2025 two talented fellows, Albert Ingo and Martin Wahlström.

Just like Meri, also Martin has been driving karts since he was 3-5 years. In 2025 he moved to one of the leading teams in the Nordics: Talvar Racing. As a result in 2025 he secured the Finnish Championship in the IAME X30 series, but left in a bittersweet second place in the Rotax Max Challenge series, had to return there in 2026 to secure a convincing victory also there. Both of these championships Martin particularly dominated in wet conditions, often leaving everyone else 0.9 seconds (per lap) behind, and earning him the nickname Captain Wahlström.

Unlike Martin and Meri, Albert Ingo on the other hand has only competed 4 seasons. Starting in the indoor Sodi W Series, where he qualified for the world finals even, Albert has shown that it is never too late to start a new sport or craft, as long as you are prepared to invest in practicing. Already at the end of his second season Albert took a top 10 finish in the final IAME X30 series in the fast paced Mika Salo Circuit. This year he finally rose to the podium, in an international race even, bringing his kart to the finish line third, in the chaotic but also very entertaining rainy race in Kouvola, Rotax Nordic Challenge 2026 race.

Talvar friends Albert and Martin also scored a memorable P1-P2 in the Vihti race this season, where Albert had dominated the practice on his home track, but Martin – not exactly a stranger to Vihti himself – was eventually able to secure pole position and subsequent victory, with a 20 millisecond advantage!

Congratulations to all three for a great season!

From a Nyrkiö point of view, just like with your software benchmarking results, we always like to apply some data science to the multitude of statistics we get from each race. One statistic that is always good to keep an eye on is the best laptime of each car. This tells you a lot about who has the best cars, whereas the finishing position of the race as a whole is a combination of many things that happen during the 20+ laps, and add up and all affect the total race time.

For example, if a group of drivers suddenly improve their lap times a lot, and they are on the same team, then this is a warning sign that it’s good to keep eyes and ears open so that we don’t miss something and fall behind.

What was a total surprise to me was that a week ago I was able to separate all drivers into clusters of similar performance, using e-divisive, the same algorithm we use in Nyrkiö for change detection. This was very surprising, because lap times is not at all what it is designed to do. But e-divisive was able to group the drivers into clusters of similar lap times / speed and the end result is in my subjective opinion much better than you’d get for example from the K-Means algorithm that one would often see used in this kind of problem sriå.

The intermediate q-statistic used in e-divisive. The maximum point in each curve, is a change point.

Race finish position (y) vs laptimes (x). Albert’s lap speed is the 8th fastet,but finishiing position around p12. Presumably the thin experience shows here. In Whalström’s case it is the opposite: He struggled to build the pace on this track, but with skillful racedriving is able to at least finish way higher than the rest of his reference group.