The other day I was in Exhibition Road in London, on my way back to the underground after a meeting. The Science Museum is next door and having not visited since my children were small, I decided I’d have a quick look around.
I wandered in, thanking myself as a taxpayer for supporting the museum’s free entrance policy and set off through the ground floor past the old steam engines, which will be familiar to many generations from school visits.
Past the space display and at the rear of the first floor was a section called Making the Modern World, comprising 150 of what have been judged the most significant items dating from 1750 to 2000 in the Science Museum’s collections.
In the middle of this relatively interesting mishmash of items – which range from a V2 rocket to a Model T Ford – was a Cray-1 Supercomputer from 1976, decorated with bright red, blue and green panels.
Now it may be that I am of a certain age, but I was stunned to see this object, which I had regarded as almost a thing of legend. When I started my working life and began to learn about technology properly, the Cray was the fastest computer in the world.
It was familiar only from photos and notable for its sci-fi appearance. A tower of technology, available in various colour schemes, arranged in a C-shape, with an upholstered bench seat around the base that contained power supplies and refrigeration. Beyond that, I knew little more.
Face to face with a real Cray-1, I learned two things immediately. One: from looking at the museum exhibit where inside the C the backs are off the computer, there was a vast quantity of wire inside the Cray, which I was not really expecting.
Two: the museum’s explanatory blurb said Seymour Cray, the eponymous designer, favoured simplicity in design right down to his preferred tools of pencils and squared paper.
When I got home I looked up a few things. When the Cray-1 launched in 1976, the first model sold for $8.8m. So yes, they were expensive. That’s an equivalent of $35.5m in today’s money. Cray sold more than 80 examples of that model, which made the company a success at that time.
Cray said: “Anyone can build a fast CPU. The trick is to build a fast system.” He was a keen believer in keeping all data paths of equal length, so that data that was meant to arrive at the same time actually did so. None of the interconnects was longer than four feet.
So as well as being his initial, the C-shape of the computers processor cabinets came from keeping all the interconnects at the back of the circuit boards as close together as possible to keep the length of the wire down.
There it was in front of me – a real machine, but now 35 years old and no longer the fastest computer in the world. It felt in a way like looking at the dinosaur skeletons in the Natural History Museum next door – an echo of the past, no longer living.
I’ve always favoured simplicity and it’s pleasing to learn that Seymour Cray did the same. Gordon Bell, the man who developed the DEC VAX – youngsters, look it up – said, “Seymour built simple machines. He knew that if each step was simple, it would be fast.”
If you’re in London, pop in to the Science Museum for a look at the Cray-1. Also in the building are other historical computing artefacts including the modern-built Babbage’s Difference engine and Ernie, the Premium Bond random-number generator.
It’s an eclectic collection – and probably more fascinating to us older IT pros than to the children.
So when was the last time you came across a piece of technology that set you musing about your earlier IT career? Why not share your reminiscences?