Fujitsu Hard Drives: Toward 1 Tb per Square Inch
A laser capable of being focused to a spot on a rotating disk just 80 nanometers across is what Fujitsu needed to be able to beat competitors Toshiba and Seagate in the race toward terabit areal densities. Yesterday, Fujitsu announced they'd achieved that goal.
While Toshiba and Seagate have been in competition with one another to drive up the areal density of hard drives using new perpendicular recording technology, the scientists at Fujitsu -- whose own consumer drives have had to play catch-up recently in the quality department -- have been planning to leap-frog their competitors in one fell swoop. There's a physical maximum, they found, to how densely data can be packed even with perpendicular mechanisms.
Their objective is to overcome that physical barrier by means of a curious physical trick involving at least three devices a hard drive has never had to use thus far: a very small space heater, a virtual refrigerator just as small, and an optical reading mechanism.
You read correctly: an optical element, not a magnetic one, but not for reading the data. It's to locate the spot on the rotating disk where the heating element will work its alchemy. Up to now, Fujitsu has had two of the three elements in its back pocket. Yesterday, it announced the third: an optical element that will help future hard drives achieve areal densities greater than the 1 terabit (Tb) per square inch theoretical maximum.
The blame for the maximum limit on magnetic recording, according to Fujitsu scientists Koji Matsumoto, Akihiro Inomata, and Shin-ya Hasegawa, has to do with the size of ferromagnetic (iron) grains. You can theoretically make them as small as you want, but if you make them any smaller than they already are, they won't retain their magnetic charge over a sustained period of time. The act of writing data literally heats these grains up, which helps them retain data; but over time, as they very gradually cool, the likelihood that they'll lose their data increases as they fall victim to what Fujitsu scientists call thermal fluctuation.
As a result, you can't miniaturize the magnetic grain enough to enable read/write heads, using the current technology, to store data up to an areal density of 1 Tb/sq. in.
Fujitsu's proposed solution is extraordinary, involving changing the physical properties of the storage media temporarily, just at the point of the write operation, using specifically focused heat. At room temperature, it takes a relatively sizeable magnetic charge (coercivity) to erase the state of stored data on a disk so it can be changed. But as the temperature in the vicinity rises, the amount of charge required decreases. If you heat a material up just enough, to what's called the Curie temperature, it loses its magnetism altogether.
For the Fujitsu process to work, a heating element needs to bring the write spot on the disk up to as close to the Curie temperature as possible, though just below. As you might have guessed, what Fujitsu needs is a laser, but one which is integrated directly into the magnetic write head of the drive. Its spot size needs to be no greater than 50 nanometers (nm).
This way, the laser can heat up the precise spot on the drive where data can be stored using a minimal charge. When the spot is rapidly cooled, it then holds its charge for a theoretically long period of time.
Up to now, it's all been theory; what Fujitsu needed was what others would consider a miracle in near-field optics. Yesterday, the company's labs announced they'd achieved something at least very close to that miracle: While scientists were hoping for 50 nm x 50 nm with 2% optical efficiency, they achieved 80 nm x 60 nm, though with 17% optical efficiency.
It might just work. If it does, the terabit barrier will be broken, and the mechanism that takes manufacturers past that barrier will have a Fujitsu patent stamped all over it. If Seagate and Toshiba were wondering what's been holding Fujitsu up for so long, they just got their answer.