Seagate has announced it has found a way to burn down a barrier other hard-drive makers are still trying to push their way through.
Late last week Seagate demonstrated a prototype of a hard-drive that records data using heat-assisted magnetic recording (HAMR) to pack a trillion bits per square inch onto a hard disk.
HAMR is a technique invented in the mid-'50s to increase the capacity of tape, which Seagate has been touting as the Next Big Thing in storage since at least 2002.
So far HAMR has finished every race trailing Perpendicular Magnetic Recording (PMR), which became the de facto industry standard in 2006, but will reach its own physical limits at around 1 trillion bits-per-inch sometime during the next few years.
Seagate predicts the capacity of HAMR could reach 50 times that of PMR because, when it runs up against the same limitation in the material used for the surface of hard drives, it changes the material.
Storage capacity increases when the mechanism recording it can make the individual bits of data occupy less space on a magnetic disk. The smaller the bit, the more powerful the magnetic field has to be to imprint it well enough that it won't scramble or fade before being used.
Bits small enough that a trillion of them will fit within a square inch is usually considered the smallest a bit can get without becoming unstable, at least using current techniques. PMR took over from the previous standard – linear magnetic recording – to overcome LMR's density limit of 100 gigabytes per inch.
HAMR's theoretical limit, according to Seagate, could be 50 times as high because it changes the physical characteristics of the surface of a hard disk – by heating it with a laser – just as it records the data.
Seagate doesn't predict when HAMR will find its way into actual products other than vague references to "later this decade."
In this case caution is probably a good idea. Not only did Seagate push HAMR as the next-generation of storage 10 years worth of storage-generations ago, it did the same thing again in 2009, with the same level of commercial success.
Even the 1-terabit barrier, which Seagate claims just to have broken for the first time, fell in October when researchers at Singapore's Institute of Materials Research and Engineering announced a new process called bit-patterned recording (BPR) that could reach densities of 3.3 terabits per square inch.
And BPR doesn't rely on adding a whole new complex system to the process or change the response-characteristics of the disk by heating it over and over, as HAMR does.
Of course, nanopatterning isn't the only super-small data-manipulating storage technique being developed to add enough capacity to existing hard drives to make any data hoarder happy.
In January IBM announced it had developed a way to pack a full byte of data (8 bits) on 96 iron atoms, rather than the half-billion or so a typical hard drive requires.
It does require the help of a scanning tunneling microscope (STM) – a device in use in experimental physics labs for several decades but whose presence is still one of the best ways to identify a lab with a lot of grant money for research and one that hopes to get by with good math skills and plenty of pencils and paper.
It's hard to imagine a manufacturing process in which workers use electron microscopes to line up individual atoms to build components as they roll down the production line, not to mention a hard drive that needs one to read a piece of stored email.
But until you heard about nanopatterning data bits in clumps on a salted hard drive, using a laser to heat the hard drive just as a magnet imprinted a bit of data on it sounded pretty high-tech too, didn't it?
Given the length of time involved (years), I wouldn't bet too heavily that HAMR is going to be the single technology that's going to shatter the one-trillion-bits-per-inch barrier on hard drives you can pick up in Best Buy, or even in high-end servers and storage area network hardware.
When a technology is just on the edge of becoming the next big thing for long enough, something cheaper, faster and better usually cuts into line and steals all the accolades.
HAMR has all the hallmarks of one of the second-place finishers in that particular race; it was invented in the fifties, has been on the cusp of commercial practicality for a decade, just broke a major milestone of its own, but faces at least two more-advanced competitors, at least one of which is likely to be practical within the same window of time Seagate is planning for the advent of the HAMR era.
Storage isn't usually that exciting a topic, but breaking the terabit-per-inch barrier is a legitimate breakthrough.
I just doubt it's going to make much difference to either Seagate or hard-drive buyers anytime at all, not just any time this decade, as Segate hoped.