Caroline Ross, on the evidence, isn’t one to duck a challenge.
About a decade ago, fresh from postdoctoral studies at Harvard, she went to work at a California hard-disk maker. Nothing notable about such a step except that Ross, now an associate professor of materials science and engineering at MIT, was the only woman Ph.D. in a firm of several thousand.
“It was okay,” says Ross. “You can’t let those things bother you.” Anyway, it wasn’t her only encounter with potentially unsettling situations. Another was when Ross, an experienced sailor, ran one of the 2,200-mile San Francisco-Hawaii races in a 25-foot sloop.
She and her sailing partner “hit a gale the first few days, which wasn’t much fun,” she says. Though the boat was decked out for ocean racing — with styrofoam packed in to minimize sinking risks, for example — the pair still had to make their way through high winds and treacherous waves.
“At night, you could barely see the waves as they came at you,” she says. “And you had to go over each one and come down correctly so as to avoid slamming into the troughs.”
Ross is now involved in an MIT effort that’s less of a threat to life and limb but has its own challenges. With others, she’s trying to create magnetic storage devices of unprecedented versatility. If the group succeeds, one ultimate result could be the end of what is today the heart of most modern computers, the hard-disk drive — an engineering marvel that nonetheless has its drawbacks.
A hard drive’s disk, revolving at some 10,000 rpms, can store vast amounts of information. This data is written on the disk by a “read-write head,” and read back as needed.
Impressive, but there are problems. For one thing, as mechanical devices, hard drives are susceptible to breakdowns. For another, it takes time to find data on that spinning disk. This is why the stuff on hard disks gets copied to memory chips when you turn on your computer. The data on chips is accessible in nanoseconds — billionths of a second — rather than milliseconds, so making the switch prevents computer work from becoming a “hurry-up-and-wait ” nightmare.
But memory chips have drawbacks,too. “They’re made up of transistors and capacitors,” notes Ross, “so when you turn the power off, you lose the data.”
But suppose you had a chip that retained everything when the power went off? Ross is laying the scientific groundwork for just such a device.
The key to her approach is the magnet. “You have an array of magnetic structures connected with metal conductors,” explains Ross, “and if you pass currents through those conductors, you can flip the magnetization of those little elements.”
Thus, each magnet can either be turned up – representing, let’s say, a “1 “- or down, which would be a “0,” so giving the system the capability that lies at the root of every digital technology.
Problem is, these magnetic grains aren’t like those dandy bar magnets we’re familiar with — the ones that, when placed beneath a piece of paper with iron filings on top, make the iron bits form graceful curves along the lines of the bar’s field.
Ross and her co-workers envision memory chips that can store more than six billion bytes, or gigabytes, of data on a square inch. (A byte represents eight “1s” and “0s”, or as much as it takes to record a single letter in the alphabet). That means packing — and electrically linking — 50 billion magnetic grains in a postage-stamp-space.
How snug is that? If you blew up such a chip to the size of North America, there’d be more than 5,000 magnets in every square mile of that vast surface.
Minuscule magnets don’t necessarily act like their household cousins, says Ross, whose specialty is the physics of magnetism. “The magnetization vectors can spiral around, or they may go straight up and down, with a little swirl at the top and bottom.”
Then, too, you have to worry about the “flippability” of your submicroscopic magnets. If they’re too resistant, you can’t store your data; if too easily flipped, you may lose it. But Ross has come up with magnet-array designs that show good promise. And Henry Smith, professor of electrical engineering and computer science, has demonstrated it’s possible to fabricate such devices.
So the group has set the stage for computers that, rather than using hard disks and conventional memory chips, keep their information permanently stored in magnetic chips. But Ross says the first applications will probably be in devices where highly resilient data-storage systems are crucial.
“They’re likely to be used in things like satellites,” she says, “or in portable devices, like cell phones.” But even there, acceptance may be slow — partly because right now the market for memory chips of any type is abysmal.
So, hurdles remain. Still, Ross loves the research. “When I was in industry,” she notes, “I worked on all kinds of things, and designing the magnetic film that stores the data on a hard disk was the most fun.”
Moreover, when pressures mount, there’s always sailing. Though she so far hasn’t encountered circumstances as dicey as those she faced in the Pacific, Ross has several times sailed the Newport-Bermuda and Marblehead (Massachusetts)-Halifax (Nova Scotia)races.
The Atlantic, Ross notes with a wry smile, can pose its own challenges. “When it gets nasty,” she says, “it can be really nasty.”