YouTuber and internet rocket scientist Scott Manley pulled off something ridiculous and brilliant: he executed a ZX Spectrum moon landing in Kerbal Space Program using a 1982 home computer with just 48KB of memory and a 3.5 MHz CPU. The achievement feels absurd on its surface—landing a spacecraft in a modern video game using technology from the Apollo era—but the technical execution reveals why retro hardware still captures the imagination of engineers and hobbyists.
Key Takeaways
- A 1982 ZX Spectrum with 48KB memory and 3.5 MHz processor successfully controlled a lunar lander in Kerbal Space Program simulation.
- Scott Manley used Sinclair BASIC, his first programming language, to write the flight control code.
- The system relied on a ZX Interface 1 serial connection running at 9.6 kbit/s to bridge the 1980s computer and modern simulation.
- The lander had severe limitations: no guidance system and barely functional attitude control, yet still achieved a successful landing.
- The entire setup ran on a single PC using the Fuse ZX Spectrum emulator, Python bridging software, and Kerbal’s kRPC scripting tool.
How the ZX Spectrum Moon Landing Actually Worked
Manley’s ZX Spectrum moon landing wasn’t a straightforward plug-and-play affair. The original 1982 ZX Spectrum had no built-in serial port, so he added a ZX Interface 1 to enable RS-232 communication. That interface supported speeds up to 19.2 kbit/s, but Manley ran the connection at 9.6 kbit/s—slow by modern standards, yet adequate for the task. The real ingenuity lay in the software stack: Sinclair BASIC code on the Spectrum read serial port data in real-time, a Python program acted as the middleman, and Kerbal’s kRPC tool handled simulation control. The CPU spent enormous cycles just bit-banging to communicate across the serial port, leaving precious little processing power for the actual flight logic.
What emerged was a genuinely constrained system. The lander had no guidance system and barely functional attitude control, according to Manley. Flying something with those limitations required finesse—you cannot brute-force a landing when your computer is starved for speed and memory. Yet the landing succeeded, proving that constraints breed creativity.
Why This Matters Beyond the Stunt
The ZX Spectrum moon landing demonstrates something the tech industry often forgets: old hardware still teaches new lessons. Modern smartphones contain billions of transistors and process terabytes of data. The ZX Spectrum did neither, yet it controlled a spacecraft. The achievement highlights the Z80 microprocessor’s practical durability and the elegance of well-written BASIC code. Manley’s project also reveals how emulation, Python scripting, and modular tools like kRPC have democratized hardware experimentation. You do not need original 1982 hardware sitting on a shelf; you can emulate it on a PC and still push it to interesting limits.
The project resonated because it captures a peculiar intersection: nostalgia for an era when programmers understood every byte of their machine, paired with the absurdity of making that machine do something it was never designed for. A computer that shipped with 48KB of memory landing a spacecraft in a game that requires gigabytes to run is comedy and engineering in equal measure.
ZX Spectrum Moon Landing vs. Modern Spaceflight Simulation
Modern spaceflight simulators like Kerbal Space Program run on systems with thousands of times more memory, processors clocked at gigahertz speeds, and graphics cards that did not exist in the 1980s. A contemporary gaming PC effortlessly handles real-time physics, 3D rendering, and complex UI interactions simultaneously. The ZX Spectrum version of Kerbal, by contrast, existed only as BASIC code firing thrusters via serial commands—no graphics, no real-time visualization, just numbers and logic. Yet both accomplish the same fundamental goal: land a spacecraft safely. This gap between the systems reveals how much of modern computing is overhead, polish, and redundancy rather than essential function.
The Practical Limits of Retro Hardware in 2026
Manley’s setup required modern infrastructure to function. The original 1982 ZX Spectrum alone could not perform this task; it needed emulation, Python, kRPC, and a PC running Kerbal Space Program. This matters because it illustrates why retro hardware remains a curiosity rather than a practical choice. The ZX Spectrum moon landing is a proof-of-concept that demonstrates ingenuity, not a blueprint for replacing modern tools. Serial communication at 9.6 kbit/s is glacially slow by 2026 standards. A smartphone’s wireless connection is millions of times faster. Yet for a tightly scoped task—reading telemetry and firing thrusters—the slow connection proved sufficient.
Frequently Asked Questions
What is the ZX Spectrum moon landing project?
Scott Manley used a 1982 ZX Spectrum home computer to control a lunar lander in the Kerbal Space Program spaceflight simulator. The Spectrum ran BASIC code that communicated via serial port to Python software, which interfaced with Kerbal’s kRPC tool to execute the landing.
Did Scott Manley land on the real Moon with a ZX Spectrum?
No. The ZX Spectrum moon landing occurred entirely within Kerbal Space Program, a video game simulation. Manley landed on the Mun, Kerbal’s fictional moon, not Earth’s Moon. The project demonstrates retro computing capability in a controlled simulation, not actual spaceflight.
Why did the ZX Spectrum need a ZX Interface 1 for this project?
The original 1982 ZX Spectrum had no serial port, so Manley added a ZX Interface 1 to enable RS-232 communication between the emulated Spectrum and the Python bridging software. This allowed the old computer to talk to modern simulation tools.
Scott Manley’s ZX Spectrum moon landing proves that clever engineering and constraint-driven design can coax surprising results from ancient hardware. The project will not replace modern spaceflight simulators or make anyone abandon their gaming rigs. But it reminds us that computing power is not the only ingredient in solving problems—understanding your tools, writing efficient code, and accepting limitations can be just as valuable. In an era of bloatware and feature creep, there is something refreshing about a programmer choosing to work within extreme constraints and succeeding anyway.
This article was written with AI assistance and editorially reviewed.
Source: Tom's Hardware
