Moon rocket from a depopulated Earth — supply chain map

The question (reformulated)

If every other human disappeared and one person (with unlimited intellect and physical capability) had to build a Saturn V-class rocket entirely from existing terrestrial industrial infrastructure, which raw materials are non-negotiable, and which countries' mines/refineries/fabs hold them?

The interesting frame: setting aside the omnipotence handwave, this maps cleanly to the real geopolitical dependency graph of modern launch. The same answer reveals why no nation today can build orbital-class rockets unilaterally.

The supply chain

Structural & airframe

Propellants

Engine-critical refractory metals

Electronics & guidance

Heat shields & refractories

What's actually interesting

1. Four-continent floor: even ignoring fabrication, the raw inputs alone require Asia (W, REE, Mg), Africa (Co, Ta, Cr), Americas (Cu, Li, He), and Oceania (Bx, Ti). No autarkic launch program is geophysically possible.
2. Single-source chokepoints: tungsten + beryllium + helium each have one dominant national supplier (~70-85%). Lose any one and the Western launch industry stops cold within 6-12 months.
3. The hardest part isn't the metal — it's the chip: in the depopulation thought-experiment, restarting a 28nm fab solo is the wall. Fabs need ultra-pure water, photolithography, ASML scanners (Netherlands monopoly), specialty gases, and a workforce. The metals you can scrape from open-pit mines; the wafer you can't make alone in a year.
4. Why this matters today: when you read about export-control fights over rare earths or semiconductor equipment, the rocket above is the implicit hostage. Civilian space and ICBM stockpiles share most of this list.

My one-line thinkering

"The cheapest part of going to the moon is the rocket. The expensive part is the international order that makes the rocket buildable."

Cold-War comparison: how did the US and the USSR pull this off?

USA — bought the supply chain

Post-WWII, Marshall Plan + alliance bloc ≈ ownership of the global mineral graph:

• Bauxite from Suriname / Jamaica / Guinea
• Titanium from Australia
• Oil from Saudi Arabia (Aramco)
• Nickel from Canada
• Helium domestic (Texas)
• Beryllium domestic (Utah)
• Tungsten domestic + Bolivia + South Korea
• Silicon, Si fabrication, optics — entirely domestic + Bell Labs
• Operation Paperclip: ~1,600 German rocket engineers including von Braun, gifted

Apollo at peak ate ~4% of US GDP (~$250B in today's money). Money + alliance + Texas oil → solved.

USSR — built a continent for autarky

The genuinely interesting case. The Soviet union was, by deliberate Stalin/Khrushchev industrial policy, designed to be autarkic. The Urals + Siberia + the COMECON satellites covered nearly every strategic mineral:

• Titanium: VSMPO-Avisma (Urals) — still the world's largest titanium producer. Ironically, the US bought titanium from the USSR through shell companies for the SR-71 Blackbird.
• Nickel + palladium: Norilsk — world's largest deposits, still #1 globally.
• Platinum-group metals: Russia is still #1 producer worldwide.
• Tungsten: Soviet mines in Kazakhstan + RSFSR.
• Bauxite: own + Hungary.
• Iron, coal, oil, gas, copper, manganese: effectively unlimited Siberian deposits.
• Uranium: Kazakhstan (still the world's #1 producer today).
• Rare earths: USSR/Kazakhstan had reasonable reserves. The Chinese REE monopoly is a 1990s+ phenomenon — in the 1960s the Soviet bloc was fine.

The one strategic gap — helium — wasn't a problem because Soviet rocketry chose kerosene-LOX for everything. R-7 (Sputnik / Soyuz launcher, 1957–today) is all kerolox; no liquid hydrogen → minimal cryogenic helium need.

The other half: simpler engineering, on purpose

USSR didn't try to match Apollo's sophistication, they brute-forced around it:

• R-7: 7-ton LEO, kerolox throughout, modular, dirt-cheap, infinitely reproducible. It launches Soyuz capsules to this day.
• Analog/electromechanical guidance instead of integrated-circuit avionics. No ASML lithography needed because no chip needed.
• N1 (the Soviet moon rocket): 30 NK-15 engines bolted in parallel — "if metals and labour are abundant, fan out the problem." Inelegant but consistent with available resources.

And yet they didn't reach the Moon

This is the punchline. Soviet failure to land on the Moon was not a supply-chain problem:

• N1 had four launch attempts, all four exploded.
• Korolev (the chief designer) died mid-program in 1966.
• The KORD engine-management computer was crude.
• They never did an all-engines static fire test on N1.

In other words, they had the metals and the kerosene; the systems-engineering integration is what they couldn't pull off in time. After Apollo 11 the political race ended and the program quietly died.

The pattern, restated

Whoever has a continent's worth of in-house mineral supply + simpler engineering can match a globally-supplied rival on the basics — but the gap shows up at the systems level (avionics, integration, computational guidance). Modern China's strategy is exactly the Soviet playbook: lock down rare earths + tungsten + magnesium + lithium processing internally, accept that you'll lag in semiconductor tooling for a decade or two, and brute-force around it with cheap parallel manufacturing. The unsolved question — for them and historically for the USSR — is whether autarky scales to the systems level, not just the inputs.

Threads to pull later

• Could SpaceX-style vertical integration replace some of these dependencies? (Reality: not the rare earths, not the helium, not the tungsten.)
• What does a "Lunar resource bootstrap" look like once you've gotten one rocket up? In-situ Al/O₂ from regolith breaks the chain — Earth bauxite stops mattering after launch #1 of #N.
• China's deep-tech push (DUV lithography clones, ARM-replacement ISAs) — is it the systems-engineering catch-up the USSR never achieved?