Silicon Insider: The Memory Whiz Kids

March 12, 2002 -- What is the greatest high-tech accomplishment of the last half-century? The integrated circuit? The personal computer? The Internet? Mapping the human genome? Pong?

They are all extraordinary, earthshaking achievements. And yet, each had a certain inevitability.

Once engineering entrepreneur Jean Hoerni came up with the planar process for producing silicon chips, and his Fairchild compatriot Gordon Moore enunciated his "law" of semiconductor innovation, you could just about predict the next few decades of the electronics revolution.

If engineers Federico Faggin and Ted Hoff, et al, hadn't come up with the microprocessor at Intel, a similar team at Motorola would have done so soon after. With the personal computer, there were at least a dozen comparable machines on the market when Steve Wozniak built the Apple I.

As for video games, Nolan Bushnell admitted at the time that in building Atari he merely recognized before anyone else that the games he played in college on a mainframe computer could now be put into an arcade machine.

And does anyone believe that the Internet (and Netscape) was anything but an idea whose time had come?

The Miracle of Memory

This doesn't take anything away from the achievements of these remarkable men any more than does history diminish the achievements of Thomas Edison, Alexander Graham Bell or Guglielmo Marconi.

It's just that these miracles are more in the timing and the synthesis of diverse components than in any profound cognitive breakthrough. Their greatness lies in results, not invention.

But there is one, hardly noticed, arena of high technology where miracles not only occur, but have taken place with such regularity over the last five decades that they constitute one of the greatest intellectual success stories in history.

It is the storage of information, be it on punched cards, magnetic tape, Winchester disk drives, floppy disks, Zip drives, RAM or ROM, CD-ROM, and DVD.

Punch Cards, Paper Tape

I remember, as a teenager, taking a course in computer programming at NASA/Ames Research Center in Moffett Field, Calif. We — a group that included some of the future founders of the personal computer industry — were shown how to type information onto punched cards or paper tape … a process that enabled us to save and store perhaps a thousands bits of data over the span of a long day's work.

We dreamed of someday being able to use one of those awesome magnetic tape drives, which held hundreds of thousands of bits, all of which could be accessed in under a minute or two.

And we were shown photographs to drool over of the next big thing — a refrigerator-sized unit filled with giant metal platters capable of holding a million bits of data or more.

We couldn't even imagine who would need so much memory: NORAD perhaps, to track bombers and missiles, or maybe the census bureau.

Storage Seemed, Wrongly, to be Limiting Factor

Then, in the late 1970s as a cub reporter, I struggled like everyone else in Silicon Valley to fully absorb the implications of Moore's Law. One thing was certain, I was told by industry veterans, the Law could never be sustained more than another decade.

Why? Because even if semiconductor chips could keep up with doubling in power every couple years, as Moore predicted, the memory needed to support those chips never could.

After all, unlike chips, which were solid-state and digital (and thus lent themselves to rapid miniaturization), memory storage was, at least in its manufacture, analog and mechanical. Everybody knew that no mechanical device could ever jump on the express train of Moore's Law.

So the sages told me, sometime around 1985 (1990 if we were lucky) a lack of low-cost, high density memory would bring the whole digital revolution to a crashing halt.

But it never happened. It still hasn't happened. And if the latest idea in storage technology — crystal lattices, electron spin, DNA soup, quantum dots — even come close to fruition, we may not hit the memory wall in our lifetimes, or those of our children.

The Memory Crusade Marches On

How did memory ever keep up? It is a story that has never really been told.

What we do know is that several generations of incredibly creative men (and a few women), in several different professions, brought every ounce of knowledge and inventiveness to bear into stuffing just a little bit more data into a chip or onto a disk.

And this has been no great single march to glory either. Rather, it has been nudged forward along three fronts — silicon, magnetic and optical — and when one has stalled, the others have move to the fore.

There have been many heartbreaking dead-ends (i.e., bubble memory). And many, once-dominant media (cards, tape, floppies) have been abandoned after having at last reached their physical limits.

But the Great Memory Crusade marches on.

Winchester Heads, Panty Hose

I've been thinking a lot about memory lately. For one thing, there is a exhibition currently taking place here in Silicon Valley at the Museum of American Heritage celebrating the history of information storage and retrieval.

To look at those big old drives, the ones I coveted so long ago in class, is to see an embodiment of American ingenuity every bit as great as those of Eli Whitney and his cotton gin, Cyrus Hall McCormick and his mechanical reaper, or Isaac Merritt Singer and his sewing machine.

Whenever I drive down Winchester Boulevard in San Jose, I am reminded that this is where a team of IBM engineers forty years ago built the first practical hard disk drives with their innovative "Winchester" read-write heads.

The iron platters, the size of garbage can lids, were so heavy that sometimes, during their design tests, they would snap off their spindles and go careening around the lab, threatening to decapitate the researchers. To get an even coating of iron oxide on these platters, the same team cleverly learned to spray through women's pantyhose.

Can the Whiz-Kids of Memory Do It Again?

I've also been pondering the story of memory out of concern over the future. I saved every penny I had to buy a 10-meg Profile hard drive for my Apple III. My kids talk about 200-gig drives the size of silver dollars in their laptops. But we've only just begun.

The next generation of the Net, Internet II, will require staggering amounts of information storage — not just in immense disk drive "farms" with capacities that require whole new Greek prefixes to measure, but also, if we are to download television shows and first-run movies at home, terabytes in our laptops and car dashboards.

Can the whiz-kids of memory do it again? I have no doubt. They may not have Moore's Law to guide them, but neither does it limit them. When you are working on ingenuity alone, there are no limits. That's why they will pull off yet one more miracle, and do it again five years from now.

In the meantime, our task is to remember … and to celebrate what the Memory Men have accomplished.

Michael S. Malone, once called “the Boswell of Silicon Valley,” is editor-at-large of Forbes ASAP magazine. His work as the nation’s first daily high-tech reporter at the San Jose Mercury-News sparked the writing of his critically acclaimed The Big Score: The Billion Dollar Story of Silicon Valley, which went on to become a public TV series. He has written several other highly praised business books and a novel about Silicon Valley, where he was raised. For more, go to Forbes.com. And you can talk back to Silicon Insider via e-mail.