How a 150-Foot Cable Limit Shaped Modern Computing

The 150-Foot Rule That Changed Everything

You might think the most important decisions in computer history were about chips or software. But one of the most impactful was about something far more mundane: how long a cable could be.

In the early 1970s, engineers at Xerox PARC were designing the Altos—the first true personal computer. It wasn't just a machine; it was the centerpiece of a revolutionary office system. The plan was simple: wire up an entire floor with these desktop computers, all connected to a shared laser printer.

But they hit a problem. The printer was noisy. Everyone hated it. So the obvious fix was to move the printer to its own room. That's when they ran into physics.

The Ethernet Constraint

The engineers chose coaxial cable for their network. It was cheap, reliable, and well-understood. But coax has a limit: the signal degrades over distance. After extensive testing, they found the maximum reliable length was exactly 150 meters. Any longer, and data started getting corrupted.

This wasn't just a number on a spec sheet. It defined the shape of Xerox's entire office network. All computers had to sit within a 150-meter radius of the printer. This forced a design decision: instead of building a single massive computer for the whole building (the prevailing mainframe model), they built many small ones, each powerful enough to work independently but connected in a tight local cluster.

Why Not Just Use Repeaters?

You might think, "Why not add signal boosters?" Well, they did. But repeaters introduced latency, and the Altos needed real-time responsiveness. The network was designed around a concept called "carrier sense multiple access with collision detection" (CSMA/CD). The idea was simple: listen before talking, and if two computers broadcast at once, back off and retry.

The distance limit directly shaped how this protocol worked. The entire network had to know when a collision occurred within a specific time window—the "collision detection time." That window was calculated based on the round-trip time of a signal traveling 150 meters. Any longer, and computers couldn't properly coordinate.

This became the famous "slot time" in Ethernet—the fundamental timing unit that all network devices used to avoid talking over each other.

The Domino Effect on Computer Architecture

The 150-meter limit didn't just affect networking. It cascaded into computer design itself.

Because the network was local and fast, each Altos could be less powerful. It didn't need its own hard drive—the network could fetch files from a shared server. It didn't need a huge memory—the network could page in data on demand. This is the birth of the "thin client" model, where the network is the computer.

But more subtly, it influenced processor design. The Altos used a custom CPU that prioritized low latency over raw compute power. It needed to respond to network events quickly, not crunch numbers for hours. This led to the first "interrupt-driven" architecture in personal computing—the idea that the CPU would pause what it was doing to handle network traffic immediately.

Today, every modern operating system uses interrupts. That lineage traces back to the decision about cable length.

The Mainframe Killers

IBM's mainframes were centralized giants. They cost millions, filled entire rooms, and forced everyone to share time on a single machine. The Xerox office system, by contrast, was decentralized. Each person had their own computer, connected by a short cable to a shared printer.

The 150-meter limit meant you couldn't wire a skyscraper. But you could wire a floor. This changed the economics of computing. Instead of one machine for hundreds of users, you had hundreds of machines for hundreds of users. Each one cheaper, faster, and more responsive.

This is the architectural DNA of the modern internet. The web works because local networks (like Ethernet) are fast, reliable, and simple. That simplicity was forced by a physical constraint: you can only push electrons so far down a wire without them getting fuzzy.

What We Lost

The 150-meter limit wasn't all good. It created a fragile topology. If the printer room cable snapped, the whole floor couldn't print. It also made long-distance networking impossible within a single building—you had to use phone lines or dedicated fiber, which were slower and expensive.

These problems shaped the next generation of networking. Token Ring, FDDI, and eventually fiber optics all emerged from trying to break the cable-length barrier. But ironically, the simple, cheap Ethernet that came from the 150-meter constraint proved more resilient. It scaled by building smaller rings, not longer cables.

The Legacy

Today, Ethernet runs on copper or fiber, and the physical length limit is a mile or more. But the protocol still uses the same slot time calculations from 1973. The timing is adjusted for longer distances, but the core collision detection logic is unchanged.

Meanwhile, the entire philosophy of distributed computing—your phone, laptop, cloud servers, all talking over a local network—grew from that cafeteria floor in Palo Alto where engineers decided a noisy printer had to move.

Next time you plug in an Ethernet cable, remember: you're not just connecting to a network. You're connecting to a decision about 150 meters of coaxial cable, a noisy printer, and the question of what a computer should be. The answer was smaller, faster, and closer to you. And it all started with a ruler.