Besides designing a new spaceship and booster,
SpaceX's engineers are busy working on a new, more efficient engine to power them.
Called Raptor, it is designed to burn super-cold methane rather than the kerosene that fuels the company's current Merlin engines.
The Starship will sport six Raptor engines. But each Super Heavy booster will need somewhere between 24 and 37.
The result will be a plumber's nightmare.
Mr Musk has said, perhaps optimistically,
that a Starship prototype might be ready for a test flight all the way to orbit (albeit without its booster stage) within six months.
That would be of a piece with its frenetic development schedule.
The rocket-building industry is used to generous government contracts that are about job creation as much as rocket creation.
SpaceX has brought a different sensibility, closer to the rapid-fire development practices of the software industry.
The Starship prototype, for instance, was welded together in a matter of months.
It was built out in the open, rather than in a carefully controlled factory environment.
The firm has two teams competing against each other to produce the best design.
Ideas are tested quickly, taken forward if they work, and scrapped if they do not.
The Starship was originally to be made of a carbon-fibre composite. But the company soon abandoned that idea, destroying its production tools.
Steel, noted Mr Musk, has a higher melting point than carbon fibre, making re-entry easier. It is also an order of magnitude cheaper.
To see the contrast, look at the Space Launch System (SLS),
another super-heavy rocket designed to ferry astronauts to the Moon and Mars, but which is being built by NASA, America's space agency.
The SLS has had around $14bn of taxpayers' money since it was authorised in 2011—
and that understates the true cost, since the SLS incorporates technology from old, abandoned rocket projects.
It is due to make its first flight in 2020, though NASA has hinted that date may slip.
Mr Musk claims that less than 5% of SpaceX's resources are dedicated to Starship. Yet it stands a good chance of beating the SLS into orbit.