Graphene Could Radically Improve Hard Drives, If We Could Only Make the Stuff
It’s been a while since we talked about graphene. While the substance drove a wave of innovation and discovery some years ago, practical problems with manufacturing have kept it out of most products. Researchers have discovered that it could be phenomenal for hard drives, but we’ll have to solve a lot of problems with the material before that can happen.
Hard drive manufacturers have turned to increasingly esoteric methods of improving performance and capacities. Helium-filled hard drives are common in data centers and enterprise markets. Using helium inside these drives reduced their operating temperatures and allowed manufacturers to fit more platters into the same physical space, partly by reducing the space between each platter. More recently, drive manufacturers have adopted support for technologies like HAMR (Heat Assisted Magnetic Recording) and MAMR (Microwave Assisted Magnetic Recording).
In this experiment, researchers replaced the carbon-based overcoat (COCs) on the drive platters with one to four layers of graphene. Graphene is incredibly thin, protects against corrosion, and is stable in a range of temperatures. These are ideal qualities given the operating environment inside a hard drive.
According to the team, those layers of graphene reduced corrosion by 2.5x and boosted drive areal densities by a factor of 10, to more than 10TB/s of capacity.
“Demonstrating that graphene can serve as protective coating for conventional hard disk drives and that it is able to withstand HAMR conditions is a very important result. This will further push the development of novel high areal density hard disk drives,” said Dr. Anna Ott from the Cambridge Graphene Center, one of the co-authors of this study.
Improving drive capacities by a factor of 10 while decreasing wear from corrosion sounds like a win/win for all concerned, but there’s an interesting wrinkle in all this. Seagate discussed it a few weeks ago when it launched its dual actuator drives.
As drive capacities have become larger, it’s become increasingly difficult for drive manufacturers to guarantee high IOPs (input/output operations per second). Single-actuator drives were no longer capable, long-term, of keeping HDDs in the performance range they need to occupy for QoS. Solutions like short-stroking a drive improve IOPs at the cost of capacity.
Increasing drive storage by a factor of 10x sounds like it would require additional performance improvements to take advantage of. Hard drive manufacturers moved away from 10K and 15K RPM drives because they were loud, noisy, and consumed a great deal of power. They also tended to harm the performance of nearby drives due to vibration. Vibration can slow the performance of a hard drive or force it to halt altogether.
The even larger problem, of course, is the difficulty of producing graphene itself. We’ve yet to invent a method of producing it reliably. As Harvard notes: “Currently, the best way to produce graphene is to take graphite, the kind you might use in a pencil, remove a layer with regular office tape, and then separate off the single-layer graphene sheets.”
“Remove a layer with regular office tape” is the kind of charming, whimsical production method that scales fine at the small scale and abominably when you want to integrate the stuff into a production environment. Graphene has been adopted in some other products — semiconductors haven’t been its primary market — but it’s very hard to work with at the scale required for factory integration. A few years ago, tests revealed that roughly half the graphene for sale was actually ordinary graphite, which did little to help adoption or spur research. Hopefully, that situation has been resolved.