The 2026 power unit is a 50/50 hybrid: ~400 kW from internal combustion, 350 kW from electric. Total system output ~750 kW (1,006 hp).
Configuration: 1.6-liter V6, 90° bank angle, 15,000 RPM limit. Single turbocharger with wastegate control. Direct injection with variable valve timing. Bore and stroke are constrained by regulation to prevent extreme short-stroke designs that would push the power band too high for the tire compounds to exploit.
Output: Approximately 400 kW (~536 hp) from the ICE alone. The thermal efficiency of these engines approaches 50% — roughly double that of a road car engine. Every percentage point of thermal efficiency gained is worth several kW at the crankshaft.
Sustainable fuel: 100% non-fossil-origin fuel is mandated from race one. The fuel must be a drop-in replacement — same energy density targets, same combustion characteristics. Fuel suppliers (Aramco, Shell, Petronas, ExxonMobil, Castrol) have spent years developing synthetic and bio-derived blends that meet the FIA specification. The stoichiometric air-fuel ratio changes slightly with different fuel chemistries, requiring recalibration of the entire injection and combustion mapping.
Specification: 350 kW motor-generator unit coupled to the crankshaft. This is 3× the power of the 2025 unit (120 kW). The MGU-K operates bidirectionally: under braking, it acts as a generator, converting kinetic energy into electrical energy stored in the battery. Under acceleration, it draws from the battery to deliver 350 kW of additional drive torque.
Energy recovery: Under braking, the MGU-K can recover up to 8.5 MJ per lap (regulated limit). The rear axle braking force is split between the hydraulic brakes and the MGU-K generator torque. The brake-by-wire system manages this split in real time, maintaining consistent pedal feel for the driver while maximizing energy recovery.
Specification: Lithium-ion cells, approximately 20 kg, 9 MJ per lap usable energy. System voltage exceeds 800V — in the range of modern EV hypercar platforms. The battery is not a single monolithic pack; it is an array of cells organized into modules with individual cell monitoring, balancing circuits, and thermal management channels.
Thermal management: Li-ion cells operate in a narrow temperature window — typically 25–45°C for optimal performance and longevity. Below this range, internal resistance rises and power delivery drops. Above it, thermal runaway risk increases. Liquid cooling circuits run through the pack, regulated by the ECU based on cell temperature, current draw, and ambient conditions. In hot climates (Bahrain, Qatar, Las Vegas), battery cooling is a limiting factor on deployment strategy.
Cell chemistry: Teams and suppliers are exploring high-nickel NMC (nickel-manganese-cobalt) and NCA (nickel-cobalt-aluminum) cathode chemistries for energy density. Silicon-carbon composite anodes are under development for higher charge acceptance rates. The FIA mandates specific safety standards for cell behavior under crash loads, including nail penetration and crush tests.
Configuration: Single turbocharger with wastegate control. The turbine sits in the exhaust stream and drives a compressor via a common shaft. Without the MGU-H to motor the compressor, the turbo must spool from exhaust enthalpy alone. At low RPM and during gear changes, there is a finite time delay before boost pressure reaches target — classic turbo lag.
Lag mitigation: Teams will employ several strategies to manage lag. Wastegate control can maintain turbine speed during lift-off by keeping some exhaust flow through the turbine. Anti-lag systems (ALS) inject fuel into the exhaust manifold during deceleration, sustaining combustion upstream of the turbine to keep it spinning. Variable turbine geometry (VTG), if permitted, adjusts the effective turbine nozzle area to match flow conditions across the RPM range. The choice of compressor map — trading peak efficiency for wider operating range — becomes a key design decision.
For the first time since 1994, F1 cars have actively adjustable aerodynamic surfaces. The old DRS button is replaced by a fully automated, continuously variable system.
When: Cornering, braking zones, low-speed sections.
Front wing: Flap angle at maximum incidence. Multi-element wing generates peak downforce. CL maximized.
Rear wing: Full-span flap closed. High-pressure upper surface, low-pressure lower surface. Maximum induced drag accepted as trade-off for grip.
Downforce: Peak aero load ~5,000 N at 250 km/h. The car could theoretically drive inverted on a ceiling above ~200 km/h.
When: Straights, DRS-equivalent zones, high-speed sections.
Front wing: Flap angle reduced to minimum. Pressure differential decreased. Drag reduced significantly.
Rear wing: Flap opens fully — far beyond the old DRS aperture. Near-stall angle on main plane. CD minimized.
Top speed gain: Estimated 15–20 km/h increase over Z-mode at the same power output. Larger delta than DRS ever provided.
Actuators: Front wing flap adjustment is driven by hydraulic or electric actuators embedded in the endplates. Transition time between modes is approximately 0.5 seconds — fast enough for the braking zone into a corner, but not instantaneous. The structural loads on the flap hinges are enormous: aerodynamic pressure at 300 km/h applies hundreds of kilograms of force per element.
Control system: The wing angle is not driver-controlled. The FIA-mandated standard ECU determines wing position based on car speed, track position (GPS), and proximity to other cars. The logic is deterministic and identical for all teams — no team can gain an advantage through control software. The driver sees the mode state on the steering wheel display but cannot override it.
Failure modes: If a wing actuator fails, the system defaults to a safe state (high-downforce Z-mode) to prevent sudden loss of grip. Redundant position sensors on each flap element confirm that commanded and actual angles match within tolerance. A mismatch triggers an automatic safety flag to race control.
Flow physics: An F1 car generates downforce through three mechanisms: the wing surfaces (upper/lower pressure differential via Bernoulli), the underbody (ground effect via Venturi acceleration), and body surfaces (diffuser, bargeboards, floor edges). In Z-mode, the front wing operates at high angle of attack, creating a strong low-pressure region on the suction surface. The rear wing does the same. The underbody tunnels accelerate air beneath the car, reducing static pressure and pulling the car toward the ground.
Transition aerodynamics: When the car switches from Z-mode to X-mode, the pressure field over the entire car changes. The front wing sheds load first (0.5s transition), which shifts the aerodynamic center of pressure rearward momentarily. The rear wing opens shortly after. During transition, the car experiences a transient pitch moment that the suspension must absorb. Teams will spend significant CFD and wind tunnel time optimizing the transition sequence to minimize pitch instability.
The structural engineering behind a 768 kg machine that withstands 5G cornering loads and 350 km/h impacts.
Construction: Carbon fiber composite survival cell. The monocoque is a semi-monocoque structure — a stressed-skin design where the outer shell carries structural loads. Layers of pre-impregnated carbon fiber fabric (prepreg) are laid up over an aramid (Kevlar) honeycomb core, then cured in an autoclave at ~130°C and 6–7 bar pressure. The result is a structure with an extraordinary strength-to-weight ratio: specific tensile strength exceeding 2,500 kN·m/kg.
Crash testing: The monocoque must pass a battery of FIA crash tests before it can race. These include frontal impact (nose cone absorbs energy, deceleration must not exceed specified g-levels), side impact (energy-absorbing panels protect the driver), rear impact (gearbox casing and rear crash structure), and a static roll-hoop load test. The survival cell must maintain its structural integrity — the driver's feet, legs, and torso must remain within the deformation envelope.
Material: Grade 5 titanium (Ti-6Al-4V). Mass: 9 kg. The Halo is a single-piece forging — no welds, no joints. Three-point mounting: two rear attachments to the roll hoop structure, one front attachment to the chassis centerline.
Load rating: 125 kN static load — equivalent to supporting the weight of a London double-decker bus (12.7 tonnes). The FIA test applies this load quasi-statically at the top of the Halo for 5 seconds. Deflection must remain within specified limits. The structure must also withstand a 50 kN lateral load.
Configuration: Push-rod front, pull-rod rear (typical, team-dependent). Carbon fiber wishbones form the primary locating members. The spring, damper, and anti-roll bar assemblies are packaged inside the monocoque or gearbox casing to minimize aerodynamic disruption.
Third element: In addition to the conventional spring-damper at each corner, teams run a third spring/damper (and often an inerter) to control heave — the vertical motion of the entire car. Heave control is critical for ground-effect cars because ride height directly determines underbody downforce. A few millimeters of ride height change can shift the aerodynamic balance dramatically.
Design: The underbody features shaped tunnels with a Venturi profile — the cross-sectional area narrows, accelerating airflow and reducing static pressure beneath the car. This pressure differential between the upper and lower surfaces generates approximately 40% of total downforce. The diffuser at the rear expands the airflow back to freestream conditions, recovering pressure and maintaining attached flow.
Ride height sensitivity: Ground-effect downforce is extremely sensitive to ride height. Too low: the tunnels choke, flow separates, downforce collapses suddenly (porpoising). Too high: the Venturi effect weakens, downforce drops gradually. The optimal operating window is narrow — a few millimeters — and the suspension must maintain the floor within this window over bumps, kerbs, fuel load changes, and tire wear. This is why heave control and third-element tuning are among the most closely guarded setup parameters.
| Material | Application | Key Property | Density (kg/m³) |
|---|---|---|---|
| Carbon Fiber Composite | Monocoque, bodywork, wishbones, floor | Specific strength: 2,500+ kN·m/kg | ~1,600 |
| Ti-6Al-4V (Grade 5) | Halo, fasteners, suspension uprights | Yield strength: 880 MPa, fatigue life | 4,430 |
| Aluminum (7075-T6) | Gearbox casing, radiator cores, brackets | Machinability, thermal conductivity | 2,810 |
| Magnesium (AZ91D) | Wheel rims, gearbox internals | Lightest structural metal | 1,810 |
| Steel (maraging) | Gears, shafts, springs | Hardness, wear resistance | 8,000 |
| Inconel 718 | Exhaust system, turbo housing | Creep resistance at 700°C+ | 8,190 |
| Aramid (Kevlar) | Honeycomb core, impact panels | Energy absorption, shear strength | ~1,440 |
| Tungsten | Ballast | Maximum density for CG tuning | 19,250 |
A rolling data center: 300+ sensors, ~1,000 telemetry channels, 1.5 TB per race weekend.
Hardware: FIA-mandated McLaren Applied Technologies standard ECU (TAG-320B or successor). Identical hardware for all 22 cars. The ECU runs all engine management (injection timing, ignition, boost control), energy management (MGU-K deployment/recovery scheduling), active aero control, brake-by-wire, clutch control, and data logging. Teams write their own software within FIA-defined parameters — the hardware is spec, but the control strategies are proprietary.
Telemetry: Approximately 1,000 channels of data are logged at rates from 1 Hz (ambient temperature) to 1 kHz+ (vibration, knock detection). Channels include engine parameters (RPM, manifold pressure, exhaust temperature per cylinder, oil pressure, water temperature), chassis parameters (ride height, suspension travel, steering angle, brake pressure, wheel speed), and electrical parameters (battery voltage, current, cell temperatures, MGU-K torque).
Volume: Each car generates approximately 1.5 TB of data per race weekend across practice, qualifying, and race. This includes raw sensor data, derived channels (calculated by the ECU in real time), video feeds (onboard cameras are separate from telemetry), and GPS positioning.
Live link: A subset of telemetry is transmitted in real time from the car to the pit wall via a dedicated radio link, and from the pit wall to the team's factory via satellite uplink. Bandwidth is limited by regulation — teams cannot stream all 1,000 channels live. They must choose which parameters to prioritize for real-time monitoring. The factory can run simulations against live data but cannot send setup changes back to the car during parc fermé.
Interface: 20+ buttons, multiple rotary dials, clutch paddles, shift paddles, and an integrated LED/LCD display. The steering wheel is the driver's sole interface with the car's systems. Functions accessible from the wheel include: brake bias adjustment (front/rear percentage), differential settings (entry/mid/exit), engine mode (power/fuel saving), MGU-K deployment strategy, radio, drink, pit limiter, and neutral.
Display: Multi-color LED array plus an LCD screen showing lap time, sector deltas, tire temperatures, energy state, gear position, and DRS/aero mode status. The display layout is driver-customizable — some drivers prefer a minimal view (gear number and delta only), others want full telemetry. At 300 km/h, the driver has approximately 0.2 seconds of glance time available for the display.
Cost: A single steering wheel assembly costs approximately $50,000–$100,000. Teams typically build 4–6 per season per driver, with spares. Each is bespoke — grip shape, button placement, and paddle geometry are tailored to individual driver hand anatomy and preference.
| Tool | Regulation | Constraint |
|---|---|---|
| CFD (Computational Fluid Dynamics) | Energy-capped | 500 MW·hr per 8-week Aerodynamic Testing Period (ATP). Teams allocate compute budget across RANS, LES, and DES simulations. Higher-fidelity simulations burn through the allocation faster. |
| Wind Tunnel | Run-limited, scaled by position | 80 runs/week for the lowest-placed constructor (P10), scaled down to 56 runs/week for the champion (P1). 60% scale model. Wind-on hours and occupancy hours are also capped. |
| Driver-in-Loop Simulator | Unrestricted (within cost cap) | No regulatory limit on simulator time. Teams run multi-day simulator programs for setup optimization, track learning, race strategy validation, and development correlation. The simulator is the one tool where more money directly buys more time. |
Pirelli is the sole supplier. The tire is the only component that touches the track — and the single largest variable in race strategy.
| Compound | Type | Grip Level | Durability | Use Case |
|---|---|---|---|---|
| C1 | Hard | Lowest | Highest | High-degradation circuits, long stints. Silverstone, Barcelona. |
| C2 | Medium-Hard | Low-Medium | High | Versatile hard option. Often the race tire at demanding circuits. |
| C3 | Medium | Medium | Medium | Baseline compound. Allocated to most circuits. Balanced grip/life. |
| C4 | Medium-Soft | Medium-High | Low-Medium | Qualifying compound at most circuits. Short-stint race option. |
| C5 | Soft | Highest | Lowest | Monaco, Singapore — slow circuits where degradation is low. Peak grip for qualifying. |
| Intermediate | Wet (light) | — | — | Damp/drying track. Tread pattern disperses standing water up to ~25 L/s per tire. |
| Full Wet | Wet (heavy) | — | — | Standing water. Disperses up to ~65 L/s per tire. Raised tread depth for aquaplaning resistance. |
Rim size: 18-inch (retained from 2022 regulations). The move from 13-inch to 18-inch rims in 2022 reduced tire sidewall height, decreasing sidewall deflection and making handling more predictable. The tire is no longer a significant spring element in the suspension — that role has shifted entirely to the mechanical suspension.
Width reduction: Front tires narrowed from 305 mm to 280 mm. Rear tires narrowed from 405 mm to 375 mm. This reduces the mechanical grip footprint, which is deliberate — with active aerodynamics providing more total downforce, Pirelli and the FIA want to rebalance the grip sources toward aero. Less mechanical grip means the active aero delta becomes more meaningful for lap time.
Tire blankets: Preheating temperature targets are being reduced for 2026, continuing the phase-down toward eventual elimination. The current target is ~70°C blanket temperature (down from 100°C in earlier eras). The out-lap becomes more critical as drivers must manage cold tires for longer. Tire compound design must account for a wider operating temperature range — the rubber must deliver acceptable grip from 50°C (cold) through 110°C (peak operating).
Computational race strategy: Monte Carlo simulation, tire degradation models, and real-time optimization under uncertainty.
The problem: Given a set of tire compounds, a fuel load, a degradation model, traffic predictions, and a probability distribution over safety car deployments, find the pit stop strategy (timing, compound selection) that minimizes total race time. This is a stochastic optimization problem with discrete decisions (when to stop, which compound) and continuous state variables (tire age, fuel load, track position).
Tire degradation model: Each compound degrades as a function of laps driven, fuel load (heavier car = more energy through the tire), track temperature, driving style, and setup. Teams model degradation as a time penalty per lap that increases nonlinearly with tire age. The "cliff" — where degradation accelerates sharply — is the most important parameter to identify, and it shifts with conditions.
Safety car probability: Historical data shows that a safety car deployment occurs in approximately 60–70% of races, with a roughly uniform probability per lap (Poisson process). The expected number of safety car laps per race is ~4. This probability must be folded into the strategy: a safety car provides a "free" pit stop (reduced time loss), so strategies that keep optionality for late stops can exploit this. The probabilistic value of waiting is non-trivial and depends on current track position and tire state.
Target: Less than 2.0 seconds stationary. The current record is 1.80s (Red Bull, 2019). The pit crew consists of 20+ members with defined roles: 3 per wheel (gun operator, tire off, tire on), 2 front jack, 2 rear jack, 1 front wing adjustment, 1 lollipop/traffic. Each action is practiced thousands of times per season.
Wheel guns: Pneumatic impact guns operating at approximately 3,000 RPM. The wheel nut is a single central-lock design — one nut per wheel, captive on the axle. The gun operator must hit the nut, spin it off, the old tire is removed, the new tire is presented, the nut is spun on and torqued, all within the sub-2-second window. Reaction time from car stop to gun engagement is typically 0.15–0.20s.
Monte Carlo race simulation: Teams run thousands of race simulations in real time during the race itself. Each simulation samples from probability distributions over safety car timing, tire degradation rates, competitor strategy, and weather. The strategy engine evaluates the expected race finishing position (not just time — position matters because points are nonlinear) for each candidate strategy and recommends the option with the highest expected value. As the race progresses and uncertainty resolves, the simulation narrows and recommendations sharpen.
Machine learning applications: ML is used extensively behind the scenes, though rarely discussed publicly. Key applications include: