Magnetic Bubble Memories

The thin-film sensor designed for hard disks first saw the light of day as a magnetic bubble detector .   Magnetic bubble memories were the hot technology of the 1970s, and the basic magneto-resistive sensor patent (a star of IBM's patent portfolio) was shared by inventors from IBM's magnetic recording and magnetic bubble research groups. Here's the paper.
      The sensor was next used in a 52-bit magnetic bubble memory chip delivered to NASA by IBM in 1971 as a feasibility study for replacing space-borne tape recorders. The next seven years saw an amazing ten-thousandfold increase in magnetic bubble memory density. Great strides were made in fabricating bubble materials and in understanding their physics. International conferences were devoted to magnetic bubble memory and logic. (We even dreamed of bubble displays in those heady days.) The project grew to several hundred IBMers, with a pilot manufacturing line in San Jose.
      The thin-film sensor, meanwhile, enjoyed success in hard drives even with the small 2% resistance change of those early days. The invention of "giant magneto resistance" boosted the signal tenfold or more, and ushered in the era of iPods, Microdrives, and laptop hard drives that hold as much as a room-filling 1980s mainframe.
      So what happened to magnetic bubble memories? They had the advantage of non-volatility, that is, they kept their information even if the power was removed, unlike semiconductor memories, and they had no moving parts, unlike hard disks. But too much time was spent debating which advantage to exploit, and in the meanwhile the entrenched technologies made sufficient progress (due in part to the work described by Krongelb & Romankiw) to make the question moot. The 52-bit chip delivered to NASA would be the only magnetic bubble device shipped by IBM.
      Was the magnetic bubble project a failure? Maybe not. When world chess champion Garry Kasparov was defeated by Deep Blue, a highly modified IBM SP2 supercomputer, it was credited with bringing hundreds of millions of dollars into IBM, even though we didn't sell any more actual Deep Blues than Ford dealers sell NASCAR racers, which is to say, none.
      In its heyday, the magnetic bubble project entertained at least one Fortune 500 company delegation per week, brought by the account representative to see what was coming from those clever researchers at IBM. We grumbled about the distraction from "important" work. But we often heard that the customer had been particularly impressed by our very visual technology. Who knows how much that was worth?