Happy 50th to the Hard Drive!
Today's Wall Street Journal had a short piece (not online) by their junior tech guy Lee Gomes on the 50th birthday of the humble hard drive. Like most computing advances of that era, the hard drive came out of IBM research, a larger-scale extension of the fixed-head magnetic disks that served as main memory before the development of cost-effective solid-state RAM. In 1956, the first commercial hard drive, RAMAC (Random Access Method of Accounting and Control) weighed about a ton and stored a total of five megabytes of data at a storage density of about 2,000 bits per square inch on a 24-inch magnetic platter. The technology actually allowed more storage than that, but as Gomes' source at IBM indicates, IBM's marketing people didn't feel comfortable selling a storage unit with greater capacity. (Echoes of Esther Dyson's idiotic mantra in 1985 that "ordinary people don't need the power of a 286. It may find some use in servers." Truly, the Hedda Hopper of technology-or perhaps the Paris Hilton.)
I personally watched hard disk storage evolve starting about 1974, when as a Xerox repairman I served a number of glass-walled data centers around Chicago's financial district. I remember cleaning out a model 3100 copier when an IBM 3333 drive cranked up two or three inches from my butt, and boy, did I jump or what? The unit was bigger than an old clothes dryer, and had 200 MB removeable platter packs (left) that were dropped in from the top. (There's a nice writeup of early IBM disk storage technology here.)
Later on, I got a job as a programmer at Xerox HQ in upstate New York. We had a Xerox Alto in our department, and I actually conned my boss into ordering a removable drive pack for me so that I could play with it as time allowed. The first-gen pack stored a staggering 1 MB of data, which seemed like the entire universe at the time. (There were larger packs available for the unit, but my boss balked at their higher price.)
I think it was 1981 when I first saw a 5 MB Shugart 5" hard drive at a Xerox technology show. This boggled me even more than the Alto, because the engineers there had rigged an interface to a generic CP/M machine, implying that I could have one too-if only the price would come down.
Eventually it did. I waited until 1986, when the first affordable 20 MB drives appeared, and installed one in my IBM PC. From there it was the familiar progression of buying a new drive every couple of years with twice the capacity of the older one. In the early 1990s the jumps began to get radically bigger: I went from a 1988-vintage 40 MB drive to a 100 MB drive in 1991, and from there right to a 1 GB drive around 1994.
I now have 600 GB in my main machine, and have begun to long for the sorts of vast, solid state computer storage that SF writers have been predicting for a long time. An 8 GB Flash unit with a mechanical write protect switch on the back would be a fine thing to boot our kernels from, and if rootkits become more of a threat, somebody will have to do that. As early as 1973 I was writing stories (unpublished) about data storage in huge semiconductor blocks that were basically grown in a tank like crystals. My 1973 concept was purely chemical, but with nanotech assemblers (even relatively simple ones) we could do it in high style. What would we do with 256 yottabyte storage systems?
I don't know. But you can bet I'd think of something.