Hitachi Scrambles to Prove It Can Beat Fujitsu to 1 Terabit HDDs

In its regular prediction of future milestones at the opening of a major annual magnetic recording conference in Tokyo, Hitachi raised the bar yet again: This time, the company is promising to reach the one terabit per square inch milestone commercially by 2011. First, that might be too late to beat its competition; second, Hitachi may have to show more proof it can reach that goal even that soon.

The goal is a fairly simple one, and sometimes it has to be simple in order for engineers to be able to fathom all that's required of them bending the laws of physics to reach it. In the case of magnetic disk drive technologies, the problem is being able to store data magnetically in a space smaller than electrons themselves should typically allow for.

Last November, Fujitsu claimed it developed the technology that breaks through the barrier to reach the goal of one terabit per square inch (1 Tb/in²): an optical read head that enabled it to better spot where magnetic grains should be positioned. Then Fujitsu developed a way of using the colorization properties of oxidized aluminum to spot the locations of pits, or nanoholes.

Not one to be left in the dust in the quest for one terabit, Hitachi claimed today it has the formula to reach the magic areal density for commercial products by 2011. But part of its major technology announcement - the use of current perpendicular-to-the-plane giant magnetoresistive (CPP-GMR) heads, which Hitachi actually premiered at this same conference five years ago.

Back then, the goal was 300 Gb/in² by 2006. As Hitachi itself admitted today -- one year past that benchmark -- it's only achieved two-thirds that number.

A US patent for Hitachi's CPP-GMR technology was issued in June 2006. But even then, it wasn't the first to the line. TDK's US patent for a very similar technology was issued in 2003.

So the question becomes, what is Hitachi ready to demonstrate for the Perpendicular Magnetic Recording Conference?

Here is where the details start to get as fuzzy as a crashed Iomega drive: Today, Hitachi says it's dumping traditional tunneling magnetoresistive heads, in favor of a CPP-GMR design that reduces electrical noise and also electrical signal, but in such a way that the signal-to-noise ratio goes up anyway.

The company is also saying that to meet its interim goal of implementing CPP-GMR by 2009, it will need to reduce track width from 70 nm to 50 nm, and then further to 30 nm by 2011. But Fujitsu's optical spotting and nanoholes technology introduced last November promises a 25 nm track width within that same time frame, and Fujitsu has shown some physical proof it can meet that goal.

Meanwhile, Hitachi appears to be scraping up interim innovations such as "damage-free fine patterning and noise suppression," which may be a fine polish on the surface. But engineers attending this week's Tokyo conference will expect Hitachi to put forth more than spit and polish.