Cadillac Opens Its Engine Factory on 18 December — the Clock It Cannot Control Runs to 2031

The 18 December date sits in two places on my desk. One is the opening of General Motors' Formula One power unit factory in Charlotte, North Carolina. The other is an empty line in the same file, the one where the number of dyno cells should be written. The first date came from GM president Mark Reuss, speaking at the Autosport Business Exchange. The second line is still blank. Until it is filled, 18 December is a civil-works date — not a technical milestone.

The last race at Brands Hatch left a timing screen I never stopped reading. That weekend taught me the difference between a building and a capability. A factory door opening is an event; how many test rigs run behind that door is a separate story. Everyone has the first Cadillac story. The second one has not been written yet, and the unwritten story is where the risk lives.

Context: the eleventh team and its first year

Cadillac entered Formula One in 2026 as the grid's eleventh team, with a chassis designated MAC-26. The power unit is currently a customer Ferrari unit, an arrangement that runs through 2028. GM's own works unit is scheduled for 2029, approved by the FIA as a supplier in 2026.

The scorecard is plain: zero points in 2026, and across the last four rounds no Cadillac driver has qualified anywhere but last. Every one of the other ten teams has scored at least once. That statistic is a sentence, and the sentence is that Cadillac is not currently in a development fight with anyone — it is racing the clock.

Being new does buy something real: the new-entrant aerodynamic testing allowance. That allowance accelerates aerodynamic learning, not infrastructure build. The cost cap behaves the same way. It compresses what big teams can spend; it does not buy a new team a wind tunnel or a dynamometer.

Core: the door, then the machine

I have a rule I have not broken since 2026. That year Force India took pole at Spa and Giancarlo Fisichella finished second. The paddock called it weather. I pulled every speed-trap reading, and every speed trap is a confession the driver never meant to make. The low-drag package was fastest through the trap above 320 km/h, and the sector deltas held across all three sessions. The result was structural.

My methodology box for this piece: sample — 2026 rounds to date, one team; conditions — customer power unit, new-entrant ATR allowance; what the model cannot explain — no technical specification for the GM power unit is public. Architecture, ERS split, combustion concept, dyno status: all four are blank in the record. Any numerical claim about the unit's competitiveness would be dishonest.

A power unit factory's real capability is measured three ways — dyno cell count, combustion and test rigs, engineering headcount. Not one of those numbers appears here. What appears is a date and a sentence: "we did it in GTP." Reuss's GTP reference is legitimate; GM built a hypercar programme from nothing and won with it. But GTP has a balance of performance. Formula One does not. In F1 your mistake is not compensated, only displayed.

My read on the Charlotte facility — tag it as a condition, not a fact — is that it will be a combined GM motorsport powertrain hub serving F1 and existing GTP activity rather than an F1-dedicated site. The GTP framing points that way. If so, it is sensible. It also means the F1 programme shares capital that is already committed elsewhere.

Then the ceiling imposed by customer status. From 2026 to 2028 Cadillac runs someone else's unit. That removes any say over packaging, cooling architecture and energy deployment mapping. A customer team loses two things: lap time and setup freedom. The chassis has to be designed around an engine it does not control, because an engine it does not control will genuinely be sitting behind it. This is an established paddock pattern, and the price is paid in every sector.

Then the clock. GM's unit arrives in 2029. FIA president Mohammed Ben Sulayem has publicly committed to a V8 return by 2031, with sustainable fuel and a smaller electric share. That is not a calendar tweak; it is a partial reversal of the technical route the 2026 regulations were built around.

The arithmetic is simple, which is what makes it brutal. If V8 lands in 2031, GM gets two seasons — 2029 and 2030 — before re-homologating for a new formula. If the switch is pulled forward to 2030, it gets one. One season, for a programme costing hundreds of millions, built from zero, at risk of obsolescence before it reaches the grid. The reporting itself calls the timing suboptimal.

One clarification, because the paddock muddles it. A presidential commitment is not a rule change. 2030 or 2031 remains the single most valuable unknown in this story, and until it resolves, GM's business case rests on a condition. I write dates, but only once they are fixed.

The technical constraint nobody discusses enough is recruitment. Staffing a competitive F1 powertrain operation in Charlotte means hiring against Mercedes HPP, Ferrari, Honda HRC and Renault. Every such hire carries gardening leave, and gardening leave means a large share of the knowledge you paid for cools off in transit. Reuss's language points here — "the right people, the right engineers and the right drivers." Engineers first, drivers after. That ordering is not accidental.

On the drivers. Sergio Perez and Valtteri Bottas: two veterans, both recording their worst career results. For a new team the choice is rational — setup direction, tyre understanding, political stability all matter more than raw pace in year one. Reuss has publicly praised their patience, saying it "takes a special person." I read that as a retention signal, and retention signals are sent when outside questions are being asked.

One question hangs unanswered. Both are high-cost, high-experience drivers in a car that cannot convert experience into points. That is defensible in year one and questionable in year three. Who occupies those seats when the works unit arrives in 2029 is a question for 2028, not 2026.

Grid expansion to eleven teams carries a cost that gets under-reported. Per-team prize money dilutes. Cadillac's backmarker status is therefore not only Cadillac's cost — it is a financial burden on the other ten teams, and it will be in the room when the 2031 rules are negotiated.

Contrarian: the fault is the car, not the engine

Here is where the story and the numbers pull in different directions.

The story is: new team, customer engine, therefore at the back. The number is: slowest qualifier four rounds running. Between those two sits a third fact that gets dropped — the GM engine does not yet exist. An engine that does not exist cannot cause a lap time. The deficit being recorded now belongs to the chassis and the aerodynamics. Being slowest across four consecutive qualifying sessions is not a marginal shortfall; it says the development loop running from wind tunnel and CFD to track is not yet connected. That is an engineering-system gap, not a power unit gap.

Correlation and causation, in one line. Being last and running a customer engine are happening together, so they are being treated as cause and effect. I keep the archive so the story cannot lie to me later.

Second point, and the one I rate highest. "Logarithmic, not linear" is an honest engineering sentence; a programme built from zero does not chart as a straight line. It is also the sentence that hides the date. If you can describe a rate of development, you can describe a waypoint. Reuss gave the rate and withheld the waypoint. That gap usually means one of two things: the waypoint is not set, or it is set and expensive to say. Either way the reader's information is exactly what the sentence contains — and the sentence contains no date.

Third, the real audience for the opening. It is tempting to assume the 18 December event is for journalists. My read is three audiences: prospective sponsors, regulators, and the engineers still to be hired. For a new team, a factory photograph is a legitimacy signal, and legitimacy is the currency recruitment runs on. The date matters. Treating it as proof of technical progress is the mistake the paddock makes at every ribbon-cutting.

Takeaway: what to watch next

I count the gates before I count the crowd. Buddh drew roughly 95,000 on race day in 2026, down to about 60,000 by 2026 — and the race died of Jaypee's balance sheet, not of empty stands. I will apply the same method to Cadillac, and skip the photographs.

Three things. First, the first official facility disclosure that includes dyno cell and test rig counts. Second, the first named senior powertrain hire — which team they came from and how long their gardening leave runs. Third, whether the 2031 V8 becomes a rule or stays a presidential commitment. Those three numbers, together, will decide what the 2029 unit is. A ribbon on 18 December will not.

And one question this season cannot answer. If a manufacturer starts a power unit programme and the president of the same regulator then talks about retiring that formula two seasons later, what does the next manufacturer weighing entry trust? That question is bigger than Cadillac. It is not yet in any file.