Filling a Hall of Fame
The world wasn’t always so nonchalant about IBM’s breakthroughs. After all, magnetic storage was supposed to be dead by now, replaced by optical storage devices or some other technology. Even IBM thought so: In 1970, an IBM research scientist published a paper “proving” that the technology would never go beyond 200 megabits per square inch.
But instead of believing the company’s own experts, the team at Almaden topped one predicted limit for magnetic storage after another. They discovered engineering work-arounds for what were once thought to be hard physical limits. By 1989, the Almaden lab was packing 1 gigabit per square inch. In the following years, Almaden has upped the ante, demonstrating densities of 3, 5 and now 11.6 gigabits per square inch. The market has begun to take for granted that magnetic storage capacities will double every 18 months, following roughly the same feverish pace set by the semiconductor industry. The big news will be when IBM slows down.
Talk to Currie Munce, director of storage systems and technology at Almaden, and he’ll complain that magnetic storage scientists are the unsung heroes of the information age. Like everyone else at Almaden, Munce likes to evangelize about storage: “We’re trying to move things mechanically over millimeter distance in milliseconds and to get them to settle within tenths of microns on track,” he says. “It’s great science.”
A visit to one particular room at Almaden shows how far IBM has run with the technology. Hanging on one wall is a single rusty platter from the original 1956 drive-proudly displayed as a rock star might show off a platinum record. Back in 1956, IBM’s disk drives were refrigerator-sized boxes that held a mere 5 megabytes, on 24 platters, each 2 feet across. Today the company ships a standard PC drive that holds more than 16 gigabytes, some 3,000 times the capacity of its original product. Put the 1956 drive and the 1998 drive side by side, though, and they look the same except for scale. Dave Thompson, director of Almaden’s Advanced Magnetic Storage Recording Laboratory, says the inventors of the original disk drive could walk into his lab today and know exactly what was going on.
Storage density depends on the size of the magnetic bit: that portion of disk real estate that is given a particular magnetic orientation-north or south-to represent a binary one or zero. At the most elementary level, the goal is simple: Shrink the bits and you expand the storage capacity. But smaller bits emanate smaller magnetic fields, which in turn requires positioning the reading head-the device that senses these fields and converts them to electrical signals-closer to the spinning disk surface.
Over and over again, shrinking the technology has forced disk-drive developers to confront physical limits that first appeared to be insurmountable. For example, the head rides on a cushion of air created by the spinning disk. Conventional wisdom held that bringing the head too close to the surface would squeeze the air molecules into a space so small the supporting cushion would disappear. “There was a lot of math to back those conclusions up,” says Barry Schechtman, executive director for the National Storage Industry Consortium (NSIC), an intercompany consortium of storage manufacturers that funds basic research at several universities. The happy reality, however, was that this theory was not true. “Nature turned out to be smarter than our equations, which needed to be modified,” says Schechtman.