Scott Manley, a Scottish-American astrophysicist and programmer, has revived an old argument about computing and spaceflight. For decades, engineers have noted that modern smartwatches contain more raw processing power than the Apollo guidance computers that sent humans to the Moon. It's a useful observation, but it misses something crucial. Manley's latest project demonstrates exactly what that something is.
Manley created a project to land a spacecraft on the Moon using a ZX Spectrum to control the descent, though he had to make do with Kerbal Space Program, a spaceflight simulation game. He used an emulator for the ZX Spectrum and Interface 1, plus Python to connect the systems. The whole operation ran on a single PC, turning what could have been a novelty into something genuinely instructive.
The Sinclair Interface 1 featured an RS-232 interface that theoretically could run much faster, but Manley stuck with 9.6 kilobits per second, which proved perfectly adequate for getting a Sinclair BASIC program to respond to and direct thruster firings in the simulator. That's a serial connection speed slower than most household internet from twenty years ago. Modern smartphones handle millions of times more data per second.
Manley acknowledged that using Z80 machine code would have been quicker, but sticking with BASIC made the code more immediately accessible. After tinkering to keep the code and simulation synchronised, his lunar lander actually landed.
The constraints were real. Manley noted that the lander had no guidance system and barely had attitude control. It wasn't a complex autonomous system. But it worked. The Spectrum handled the calculations. The Interface 1 shuffled data. The BASIC code parsed numbers and managed the landing burn. The result was a successful touchdown on a virtual lunar surface.
The ZX Spectrum became one of the most important and influential computers of the 1980s, with some observers crediting it for launching the British information technology industry during a period of recession and introducing home computing to the masses, eventually selling over five million units. Manley's project doesn't diminish that legacy. Instead, it reminds us that engineering excellence has always been about working within constraints, not just having bigger numbers.
The reasonable criticism of the "more computing power than Apollo" comparison isn't that the comparison is false. It is. The criticism is that it confuses quantity with capability. Modern processors are faster, sure. But that speed is often wasted on inefficiency. A Spectrum running BASIC at 9.6 kilobits per second, controlled by someone who understands orbital mechanics and how to write lean code, proved sufficiently powerful for the task. Given enough ingenuity, older hardware doesn't always become obsolete. Sometimes it just becomes a better teacher.