Why F1 Removed the MGU-H, and What Deleting Complexity Buys You

The MGU-H recovered energy from exhaust gas by using its flow to spin a turbine, converting waste heat that would otherwise have been thrown away into usable electrical energy. It also let teams keep the turbo spinning under braking, which all but eliminated turbo lag. It was arguably the cleverest single component in a modern Formula 1 power unit, and the 2026 regulations removed it entirely.

It was not removed because it failed. It worked extremely well, and the engineering behind it was genuinely remarkable. It was removed for reasons that had nothing to do with whether it functioned.

The case against a component that worked

The MGU-H was extraordinarily difficult and expensive to develop. Recovering energy from exhaust gas means operating a precision electrical machine in one of the most hostile thermal environments on the car, at rotational speeds well beyond anything in a road vehicle, for a full race distance, reliably. The development programmes required to make that work were enormous, and the knowledge involved was closely held by the manufacturers who had already solved it.

That created a barrier to entry that compounded over time. A new manufacturer entering the sport would not simply need to build an MGU-H, they would need to build one competitive with units refined over a decade by organisations who had already spent hundreds of millions learning. The gap was not a starting handicap, it was a moving target that a newcomer would be chasing for years while being uncompetitive in public.

And its relevance outside the sport was close to zero. Formula 1's justification for expensive hybrid technology had always been that it drove development applicable to road cars. Direct injection, energy recovery under braking, thermal efficiency, all defensible on those grounds. The MGU-H was not. No manufacturer was going to put one in a production vehicle. It was a brilliant answer to a question only Formula 1 was asking.

What replaced it

The 2026 formula keeps the 1.6-litre V6 turbo and puts the recovery burden on a much more powerful MGU-K, which now delivers 350kW to the rear wheels against 120kW previously, with the regulations capping ERS-K electrical DC power at 350kW.

That is a nearly threefold increase in the significance of a component that is far more conventional. An MGU-K recovers energy under braking, which is a problem every hybrid road car already solves, at a scale and sophistication Formula 1 can push forward usefully. The energy split moved from roughly 80:20 in favour of combustion to something close to even.

The sport gave up a genuinely exceptional piece of engineering and got back accessibility, lower development cost, and a far clearer path for a new entrant.

Did it work?

This is the part where the argument stops being theoretical. Audi entered Formula 1 for the first time for 2026, taking majority control of Sauber and arriving with its own power unit from day one rather than buying someone else's. Red Bull Powertrains, with Ford as technical partner, built an engine from scratch in Milton Keynes, having never manufactured one before. Honda returned with an exclusive factory arrangement with Aston Martin.

Five manufacturers now supply the eleven teams on the grid. It is difficult to argue that lowering the technical barrier failed to attract entrants when two organisations with no prior Formula 1 engine-building history committed to doing exactly that.

Whether their engines are competitive is a separate question that the season answers, not the regulations. But the decision to attempt it at all was made possible by a rulebook that no longer required a decade of MGU-H expertise as the price of admission.

The transferable lesson

We run into this decision constantly in software, and it is almost always resolved the wrong way.

A clever abstraction that only its author fully understands is a liability from the moment they take another job. A bespoke framework that saves a fortnight during the build costs a year over the life of the product, paid in onboarding time, in bugs that nobody else can diagnose, and in the growing reluctance of anyone to touch the part of the system that is known to be subtle.

The measure of a technical decision is not how impressive it looks on the day it ships. It is how easily the next person can work with it, how quickly someone unfamiliar can become productive, and whether the thing can be handed over at all without a lengthy oral tradition attached.

We have inherited plenty of projects where a previous developer built something genuinely ingenious that the client could no longer maintain, extend or safely modify. In every case, the ingenuity was real and the outcome was still bad. The client did not need the clever thing, they needed a thing that worked and could be changed.

Removing the MGU-H made Formula 1 less clever and more durable, and it widened the field. That trade is available in most technical work, and it is worth taking far more often than it is.