Forget container ships crawling across the Pacific at 24 knots. What if we stood a 2-kilometer electromagnetic rail gun on end in California and lobbed shipping containers to Shanghai in under an hour? The physics is unforgiving, but instructive.

The ballistic problem. San Francisco to Shanghai is ~9,900 km. Ignoring Earth's rotation and treating this as a vacuum ballistic arc, the minimum-energy trajectory (a "semi-Hohmann" suborbital hop) needs a launch velocity of roughly v = √(g·R·sin(Δ)/(1+sin(Δ/2))). For a range that's ~1/4 of Earth's circumference, that works out to about 7.0 km/s at a 22.5° launch angle — essentially orbital velocity. Apex altitude: ~1,200 km. Flight time: ~42 minutes.

The gun. To reach 7,000 m/s over a 2 km barrel, the payload needs constant acceleration of a = v²/(2L) = (7000)²/(2·2000) = 12,250 m/s², or 1,250 g. A standard 30-ton shipping container becomes a slug of jelly; even ruggedized cargo (steel ingots, ore, bulk grain) survives 1,000+ g only if densely packed and shock-mounted. Human passengers: absolutely not. Electronics in potting compound: maybe.

The energy. Kinetic energy per 30-ton shot: ½·30,000·(7,000)² = 7.35 × 10¹¹ J ≈ 204 MWh. That's one Tesla Megapack's entire capacity per launch. At a modest firing rate of one shot every 10 minutes, you need 1.23 GW average draw — a full nuclear reactor dedicated to the gun. Peak power during the 0.57-second launch pulse hits 1.3 TW, requiring a capacitor bank or homopolar generator storing that 200+ MWh and dumping it in half a second.

The rails. Rail guns at even lab scale (Navy's now-shelved 32 MJ prototype) erode rails after a few dozen shots from arcing and ohmic heating. Scale to 735 GJ per shot — 23,000× the Navy gun's energy — and the rails vaporize instantly unless you use a coil gun (asynchronous linear induction motor) instead. That's the SpinLaunch/StarTram approach. Rail current at 7 km/s exit and a 1-meter armature: peak currents on order of 10 million amps. Copper's fusing current density says the rails must be ~1 m² cross-section each. That's 32 tonnes of copper per meter of barrel, times 2,000 meters — 64,000 tonnes of copper, roughly $500M just for the conductors.

Atmospheric exit. At sea level, a container-shaped projectile at Mach 20 experiences dynamic pressure of q = ½ρv² = ½·1.225·7000² = 30 MPa — 300× atmospheric pressure, enough to crush the container flat before it clears the first kilometer. Solution: launch from a mountain peak (5+ km) inside an evacuated tube, with a plasma window or fast-opening iris at the muzzle. Even then, stagnation temperature on the nose cone hits ~25,000 K. You need an ablative heat shield on every shipping container.

Landing. Shanghai doesn't want 30 tonnes of steel arriving at Mach 20. Either a matching decelerator gun (linear motor braking, absorbing that 204 MWh back into the grid — actually elegant), or a splashdown zone 200 km offshore with parachutes deployed after re-entry heating dissipates below 60 km altitude.

Verdict: for bulk, shock-tolerant, time-sensitive cargo (semiconductors, pharmaceuticals, munitions), the physics closes — but only barely, and only if you're happy dedicating a reactor and building a receiving station symmetric to the launcher.