Why Equal F1 2026 Engines Fail to Fix Mercedes Gap
McLaren and the Mercedes Power Unit in 2026: Why Equal Hardware Has Not Meant Equal Performance
In the first half of the 2026 Formula 1 World Championship, one of the most discussed technical stories has not been about a gap in equipment, but a gap in exploitation. McLaren Racing, running the same specification Mercedes AMG High Performance Powertrains power unit as the Mercedes-AMG Petronas works team, has been measurably off the pace of its supplier on power-sensitive phases of the lap.
The difference is not the result of customer parts being held back. It is the result of how dramatically the 2026 regulations have changed what it means to run a modern Formula 1 power unit, and how much advantage comes from designing the chassis and the power unit as one system.
This is not a story of favoritism. It is a story of integration, operational learning, and the new weight of electrical energy management in lap time.
The 2026 Reset Changed the Job Description
To understand McLaren’s position, you have to understand what changed for 2026. The formula removed the MGU-H, increased the electrical output of the MGU-K from roughly 120 kW to 350 kW, nearly tripled the energy storage capacity, and rewrote the fuel flow and energy deployment rules to create a car that is far more dependent on electrical energy than before.
The working split in race trim is now close to 53 percent internal combustion and 47 percent electrical. That number alone tells you how different the driving and engineering task has become. Under the previous generation, you could manage a small harvesting deficit and still recover with top-end power and aerodynamic efficiency. In 2026, if you under-harvest through Turns 7, 8 and 9, you arrive at the following straight with a lower state of charge.
You then have less boost to deploy, you spend longer at full throttle without electrical assistance, you use more fuel to compensate, and you generate more heat that you must then manage later. A small error compounds.
How Kimi Antonelli & Mercedes Mastered 2026 F1 Championship
The regulations are intentionally strict on parity. Customer teams receive identical hardware, identical baseline control electronics, and identical homologated software maps from the manufacturer. What the regulations cannot legislate is operational understanding. How you recover energy under different grip conditions, how you predict battery temperature over a stint, how you shape deployment to suit your drag level and your gear ratios, how your driver blends mechanical braking and electrical recovery at corner entry, these are learned behaviors.
They live in simulation models, in thousands of hours of dyno correlation, in driver-in-the-loop work, and in the institutional memory of the people who designed the system.
Mercedes HPP started that learning process while the hardware was still on the drawing board. Customer teams started it when the hardware arrived.
Why Identical Does Not Mean Equal
Formula 1 has operated with customer power units for decades, and in most recent seasons the gap between works and customer exploitation has been relatively small. 2026 broke that pattern because the performance differentiator moved from peak power, which is easily equalized, to energy management, which is not.
Consider what a power unit control system has to do now. In every braking zone, the system must decide how much deceleration comes from the rear axle via the MGU-K and how much comes from friction brakes, while keeping the rear axle stable, keeping the battery within its thermal window, and preparing the state of charge for the next 30 to 40 seconds of lap.
On the straight, it must decide whether to deploy aggressively early to build speed or to phase deployment to defend against drag and to protect energy for the next sequence.
Those decisions are made by calibration maps that are theoretically available to everyone, but the quality of those maps depends entirely on the quality of the models behind them. The works team arrives at the track with a model that already knows how its chassis flexes under load, how its cooling package responds to traffic, how its aero map shifts the rear grip that directly affects how much energy you can harvest without locking or instability.
A customer team arrives with a very good model of its own car, and a good but fundamentally external model of the power unit.
Early in this season that showed. Mercedes had cleaner deployment traces on long straights, with less clipping and more consistent top-end speed. McLaren, Aston Martin and Williams, all Mercedes-powered, showed more variance in state of charge from lap to lap and more lift-and-coast interventions late in stints to recover energy targets. The hardware was the same. The operating window being used was not.
The Structural Advantage of Being a Works Team
There is a practical reason why works integration matters more in 2026 than it did in 2022.
When Mercedes designs its 2026 chassis, its power unit group can ask for specific installation choices. They can place the battery and power electronics to optimize center of gravity for harvesting stability. They can shape cooling ducts around known heat rejection from the MGU-K and inverter at high-duty cycles.
They can choose gear ratios that place the MGU-K in its most efficient harvesting and deployment band for the majority of circuits, because they know the exact shape of the combustion engine’s torque curve and the electrical machine’s efficiency islands.
A customer team cannot have that conversation in the same way. McLaren designs its chassis to a set of interface requirements and estimated performance curves. It receives installation drawings, cooling targets, and operating guidelines. It gets excellent support, including resident HPP engineers, trackside assistance, and access to some simulation tools.
It does not get the ability to co-optimize the fundamental architecture through multiple loops before the car is frozen.
Andrea Stella described this early in the year in a very honest way. He said McLaren, for the first time in its recent partnership with Mercedes, felt it was on the back foot in predicting car behavior. That was not a complaint about hardware. It was an acknowledgement that the team’s own simulation had less pre-season maturity on the energy side than the works team’s.
Gear ratios are the clearest example. For 2026, teams had to nominate their ratio sets before the season with very limited opportunity to change. If you have the most accurate power delivery model, you can pick ratios that keep you in the sweet spot for both harvesting under braking and deploying out of slow corners.
If your model is even slightly off, you may end up with ratios that are a little short at high-speed tracks. That forces more shifts, more transient interruptions in harvesting, and a less stable state of charge. McLaren’s early-season straight-line deficit was partly aerodynamic, but it was also partly this. A short ratio costs you not just top speed, but energy continuity.
Where the Time Was Actually Lost
When you overlay GPS traces and energy data from the opening rounds, the pattern is consistent. McLaren did not lose three or four tenths in a single place. It lost it in a chain.
First, cornering grip. Early in the year, the MCL40 was not generating the same mid-corner rear downforce as the W17. Less rear grip means less confidence on entry, which means the driver cannot ask for as much negative torque from the MGU-K without destabilizing the rear. If you harvest less in the corner, your battery state drops.
Second, that lower state of charge means you have less electrical boost on the next straight. You compensate with more throttle time on the internal combustion engine, which increases fuel consumption and thermal load. To manage temperatures, you later have to derate deployment or adjust lift points, which costs more time.
Third, aerodynamic drag. McLaren ran a slightly higher drag level to recover some of the downforce deficit. Higher drag makes deployment less efficient, because you burn electrical energy to fight air resistance rather than to accelerate.
Marcin Budkowski Joins Cadillac as Formula 1 Team Principal
Team leadership later quantified it in a way that matches the on-track picture. Roughly 70 percent of the deficit to Mercedes at mid-season was attributed to pure cornering performance and grip, and roughly 30 percent to a combination of drag and power unit exploitation.
That 30 percent is not trivial in 2026. At 0.15 seconds per lap of straight-line loss alone, which is the kind of figure that was discussed after the early flyaway races, you are already outside the window where strategy can cover you at most tracks.
Reliability added another layer. Early battery-related management issues and a more conservative lifecycle plan meant McLaren introduced certain reliability-focused specification updates later than the works team. Those updates were not billed as performance upgrades, but in this formula, reliability is performance.
If you can run the battery hotter or cycle it deeper because you trust your new safety margins, you can run more aggressive energy maps. Until McLaren had that confidence, it had to run a little more margin.
None of this points to an underpowered customer unit. It points to a car that, for a period, created harder conditions for the power unit to do its best work.
How McLaren Rebuilt Its Approach
McLaren’s response is important because it shows how seriously top teams now take energy management as a core discipline.
Under previous regulations, power unit performance was handled by a small group within vehicle performance that worked closely with HPP. For 2026, technical director of performance Mark Temple established a dedicated power unit performance group inside McLaren.
This is not just a renamed department. It is a group of control engineers, calibration specialists, and energy strategists whose sole job is to extract lap time from the electrical system.
Technical director of engineering Neil Houldey has spoken about the learning loop that this group runs. After every session, they compare predicted versus actual harvesting by corner, by driver, by tire compound, and by fuel load. They examine how state of charge evolves not just over a lap, but over a full stint with traffic and safety car probabilities. They then rewrite deployment plans for the next session and feed lessons back into the simulator.
That simulator work has been critical. McLaren secured deeper access to Mercedes’ power unit simulation environment as the season progressed. By the time the championship reached Spa, Stella noted that the team could finally anticipate energy effects before arriving at the track, rather than discovering them on Friday. That shift from reactive to predictive is everything in this formula.
Driver technique has evolved alongside it. Lando Norris and Oscar Piastri have had to adapt their braking profiles to maximize recovery without hurting entry stability. That sounds small, but in 2026, moving the brake balance by one percent and changing how aggressively you initially apply the pedal can change harvested energy by tens of kilojoules per corner. Over a 44-lap race, that is race-defining.
The result of that work was visible well before any major car upgrade. McLaren’s energy traces became smoother, the lap-to-lap variance in state of charge narrowed, and the team’s long-run pace improved more than its single-lap pace, which is exactly what you would expect when energy management improves. The victory in Hungary was not just a chassis victory. It was a race where energy discipline, combined with improved downforce, allowed McLaren to run its planned deployment map from start to finish without forced saving phases.
What Has Closed and What Has Not
By mid-season, the pure exploitation gap, meaning the time lost because of deployment and harvesting optimization rather than aerodynamics or grip, had clearly narrowed. McLaren was no longer surprised by energy shortfalls on Fridays, and the straight-line speed gap to Mercedes had reduced in comparable wing levels.
That does not mean the gap is zero. Two residual areas remain.
First, chassis development still feeds the energy loop. Even with perfect maps, a car that generates less rear load at high speed will harvest less efficiently. McLaren’s ongoing development program aimed at improving rear floor and beam wing efficiency is therefore doing double duty. It adds downforce directly, and it creates better conditions for the MGU-K to operate.
Second, some of the deepest characterisation data still resides within HPP. Customer teams will always have slightly less visibility into the absolute limits of the battery, inverter, and MGU-K under combined thermal and electrical stress than the team that built them.
That is not a question of trust. It is a question of liability and intellectual property. Mercedes must give customers safe, reliable operating envelopes. It cannot give them every proprietary detail of how close to the edge the works team is willing to run.
McLaren has been consistent on this point. The team has emphasized a strong working relationship with HPP while also pushing, as any competitive team would, for earlier information flow and earlier access to reliability updates. That tension is inherent to the customer model and does not indicate any deliberate performance withholding.
The Bigger Picture for Formula 1
What 2026 has exposed is that the old definition of parity, same hardware, same software, equal treatment, is necessary but no longer sufficient to guarantee equal opportunity.
When the regulations place this much importance on electrical energy, the advantage of vertical integration grows. The works team does not just build a better engine. It builds a better understanding of how that engine wants to be used inside a specific chassis concept, and it builds that understanding earlier.
This is not unique to Mercedes and McLaren. Similar patterns have been observed in other power unit partnerships this season. The customer teams that have adapted fastest are those that have treated energy management as a first-order performance department, not as a calibration task for the power unit supplier to handle.
For Formula 1, the question is not whether customer teams can win under these rules. McLaren has already shown they can. The question is how much organizational investment is now required for a customer to reach the same level of exploitation as a works team, and whether the sport should consider additional standardization in simulation tools or energy model access to help close that operational gap.
F1 V8 Engines: Stefano Domenicali & Ben Sulayem Target 2030
It also reframes how we evaluate team performance. In previous eras, we could separate chassis and power unit reasonably cleanly. In 2026, you cannot. A tenth lost in aero can become two tenths lost in energy if it hurts harvesting, and a tenth gained in energy management can look like a straight-line speed upgrade when it is actually just smarter deployment.
Conclusion
The claim that McLaren has not yet fully exploited the Mercedes power unit in 2026 is correct, but only if it is understood in the right way. The data and the technical context point consistently to an integration and operational challenge, not a hardware deficit.
The 2026 power unit is fundamentally more complex to operate than its predecessor. Its performance depends on a tight loop between grip, downforce, drag, thermal management, driver input, and control strategy. Mercedes, as the manufacturer, entered the season with a deeper, more integrated model of that loop. McLaren entered with the same physical power unit but with less accumulated learning and less ability to co-optimize chassis and power unit before the season.
McLaren’s creation of a dedicated power unit performance team, its improved simulation access, its systematic analysis of harvesting and deployment corner by corner, and its parallel chassis development have steadily reduced that learning deficit. The trajectory since the early races is one of convergence, not divergence.
In that sense, this season is a case study for modern Formula 1. Regulatory parity can equalize what you are given. It cannot equalize how quickly you learn to use it.
Under a formula where energy is lap time, learning speed is performance, and McLaren’s progress shows that a well-resourced, technically agile customer team can close that gap, even if the works team will always have the first step.
For more F1 News
Follow us
Never Miss an F1 Update!
Join the AutodromeF1 Formula 1 community for real-time race commentary, paddock news, and 2026 calendar updates across your favorite platforms: