Massive reels of magnetic tape are a long-standing visual cliché. Those spinning reels are almost as popular as walls of blinking lights when a movie director needs to tell the audience, "Here is a big computer." So it could remain for decades to come because magnetic tape (with or without those photogenic reels) refuses to be displaced as a storage medium—while technical, even social, demands on our IT systems are dramatically expanding our need for low-cost archival.
The more you know about tape, the harder it is to believe that it works. As tape is wound from one reel to another, the flexible backing stretches; the magnetic coating stretches less, leading in the long run to flaking and shedding. When a magnetic field is applied to produce a record, there isnt a simple proportion between the strength of the field and the strength of the persistent change in the magnetic state. That nonlinear relationship requires well-tuned electronics, as well as a recording head that can focus the field on the smallest possible area so that data can be recorded at maximum density.
But tape refuses to knuckle under to the fact that it shouldnt work nearly as well as it does. At half a cent per megabyte of storage, tape is dirt-cheap; at 20 to 30 years of shelf life, tape satisfies our enterprise needs for even what we call long-term storage.
And tape is still getting better. In May, on the 50th anniversary of its first tape drive introduction (1.4MB on a 12-inch reel), IBM announced the achievement of storing a terabyte of data on a single 4-inch-by-5-inch cartridge—the equivalent, the company estimated, of 16 continuous days of DVD video, storing data on tape at 900M bits per square inch. In June, IBM shipped new commercial drives storing 180GB per cartridge and offering transfer rates of 42MB per second (at 3-to-1 compression).
Further reading
With "Lord of the Rings" amassing 150 terabytes of video and with corporate e-mail archives (increasingly subject to record retention requirements) and e-business transaction records in hot pursuit of comparable data volumes, we need compact and low-cost archival more than ever—and we need to avoid the assumption that tape, just because its old, is no longer a competitive solution.
Tell me what tape has done for you lately at peter_coffee@ziffdavis.com.
Research Assistant
