A 160 TB disk drive?

About every 10 years hard drive technology bumps into the laws of physics. And for the last 50 years scientists and engineers have figured out ways around it. 3-D recording may be the next bit of magic.

160 TB disk drive

Hard drives are beginning to reach the limit of current recording technology – about one terabyte per square inch. With present technology higher densities are risky because at such tiny sizes thermal fluctuations can cause bits to flip on the hard drive surface.

Regular readers of Storage Bits know that there are two promising technologies to overcome the current limitation: Heat Assisted Magnetic Recording (HAMR) and Bit Patterned Media (BPM). Magnetic bits written at higher temperatures are more stable when they cool and bit patterned media enables higher densities by enabling the use of the entire volume of the magnetic pit.

But even these heroic and not yet proven feasible for mass production technologies will only take hard drives up to about 20 TB capacity. That's where 3-D magnetic recording could save the day.

In a recent paper, authors Nissim Amos, John Butler, Beomseop Lee, Meir H. Shachar, Bing Hu, Yuan Tian, Davil Garcia, Rabee M. Ikkawi, Robert C. Haddon of the University of California Riverside, Dmitri Litvinov of the University of Houston, Jeongmin Hong and Sakhrat Khizroev of Florida International University document their success in creating an 8 3 bit 3-D magnetic cell.

With a three layer magnetic medium and some esoteric magnetic techniques - Kerr rotation - they successfully wrote and read 8 different magnetic states in a single BPM cell. The cell's three layers are different thicknesses to overcome the problem of distance from the read/write head.

Screen Shot 2013-10-14 at 8.42.34 AM
Image: courtesy of the authors

The Storage Bits take

A laboratory demonstration does not a commercially viable product make. However the fact of 3-D recording breaks a basic limitation of 2-D recording that the industry has been living with for decades. Nor is this the only exploration of 3-D recording techniques today.

What this does show is that simplistic extrapolations of semiconductor improvements - as wrongly applied to NAND flash – versus magnetic recording bit costs demonstrate nothing about the future of hard drive success.

With sufficient investment and continued refinement of the underlying technologies it should be possible to build 160 TB hard drives within 10 years. At that point, tape will be well and truly dead, as hard drives replace it.

Comments welcome, as always. When is a hard drive too big? Some thought 9 GB was too big.