RAMAC: A Lesson in Disk Drive History from Professor Hoagland
This weekend saw the tenth Vintage Computer Festival held at the Computer History Museum. It’s a nerd fest of the highest order, with a roomful of old machines, some surprisingly old like an operational 1962 LINC. (Note: The head of the LINC project was “Wes Clark,” as in retired General Wesley Clark, former NATO commander and one-time US presidential candidate. Never mind. ) I was there as a museum docent. My job for the day: open the front door for people and tell them where to find the festival. The museum likes to use all of my talents to their fullest.
However, I also got a chance to listen to a fascinating lecture by Professor Al Hoagland of Santa Clara University. Hoagland got his EE PhD from Berkeley in 1954 and he was on the IBM team that developed the world’s first computer disk drive, the 5-million-character RAMAC (random-access method of accounting and control), in 1956. Hoagland’s experiences at Berkeley working on multi-head drum memories landed him a RAMAC consulting gig and ultimately a job and a career at IBM in San Jose, which lasted nearly three decades.
Hoagland tells his RAMAC story easily. You can tell he’s told it many times. In the early 1950s, IBM decided to start recruiting technical talent from the west coast universities. Unfortunately, IBM found that it had a hard time attracting California graduates to the climate in Armonk and East Fishkill, New York. So the company decided to lease 8000 square feet at 99 Notre Dame in San Jose to start a small research facility. IBM picked San Jose because it was between two major computing markets, Los Angeles and Seattle, and was close to an established IBM punched-card plant—located in a former laundry—which could provide administrative support to the startup research facility. (The plant moved to Campbell, CA in 1960.)

The head of the place was Rey Johnson, a charismatic leader who gathered the talent and then went searching for a specific mission. That mission quickly became the development of a fast, random-access memory to replace paper vertical files and tubs of punched cards. The focus quickly became a disk drive that would be called RAMAC. RAMAC had only a pair of heads to read one disk in the 50-disk stack at a time. The paired heads rode on a little elevator to the correct level. It was apparent that a head-per-disk design would greatly improve performance and that research work started in 1955 under Al Shugart, who would lead many more revolutionary developments in the disk industry.
Before RAMAC, computers stored data on punched cards, which had been IBM’s bread-and-butter market for decades, and magnetic tape. Both storage media posed severe access problems and required a lot of manual handling to change tapes and shuffle cards from machine to machine. A disk drive promised to put a lot of data within instant reach. RAMAC was announced in 1955 and the product became available in 1956 as part of the IBM 305, the company’s first transaction-processing system. (It’s a little confusing because the RAMAC drive itself was the model 350.) The IBM 305 system was a huge success and caused IBM to build a big new manufacturing facility in south San Jose. IBM’s disk presence in San Jose was huge for the next 50 years, until IBM sold its drive manufacturing business to Hitachi.
Hoagland is now a professor at Santa Clara University. He discovered that his students had no understanding of the basics of disk drives and he set out to correct that lack of knowledge. He obtained a loan of one of the four RAMAC drives that IBM still has (just the raw drive mechanism, not the 1950’s era electronics) and his students spent two years experimenting with the drive mechanism and developing microprocessor-based control electronics to reanimate the RAMAC. So far, they’ve been able to pull 50-year-old data off of all the drive’s tracks. (See this article in Santa Clara Magazine for more info.)
Hoagland has organized the Magnetic Disk Heritage Center, which is currently co-located with the Computer History Museum in Mountain View.
Steve Leibson commented:
R. Prichard commented:
Matthew Miller, EDN.com editor-in-chief commented:















