As SSDs become increasingly inexpensive, traditional magnetic hard disk drives are becoming less attractive. Except for a few industrial storage applications, this is probably true for gamers. This remains true even with the recent upward trend in flash memory prices. But if you want maximum capacity, there is no substitute for rotating platters. That's why Seagate is demonstrating its new hard drive technology, which is expected to expand capacity to 240TB per drive over the next decade or so. [The latest heat-assisted magnetic recording (HAMR) in the state-of-the-art drive offers data density as much as twice that of traditional perpendicular magnetic recording (PMR) technology found in hard disk drives. But now a research team from Seagate Technology, NIMS, and Tohoku University has successfully demonstrated multi-level HAMR (via Tom's Hardware).

Obviously, the basic concept of multilevel magnetic recording is not new. The problem was finding the right materials. Now, a team led by Seagate has obtained two FePt-C nano-granular films separated by a Ru-C spacer layer with a cubic structure. It is obvious!

By the way, if you are a materials science PhD, you can check out the full research paper here. For the rest of you, please see the cliff notes:

By adjusting the laser power and magnetic field during writing, the FePT layers can be treated independently, thereby doubling the capacity. If this is not enough, the researchers believe that this approach could be applied to triple-layer recording, and could work for quad-layer recording as well.

If so, that would result in a 10-layer hard drive with a rough capacity of about 240 TB. This is enough capacity to install one or two of Steam's latest games.

Seagate's current BarraCuda series of desktop drives has a maximum capacity of 8GB, while enterprise-class HDDs are limited to 24GB. This means that there is a potential for 10 times the capacity of Seagate's current largest offerings.

Exactly when this technology will be available in PCs is a bit of an open question. The technology appears to be proven, but it is likely to be very costly, at least initially. Apparently, multi-level materials require a new design of read/write head.

Therefore, it is likely that this novel technology will first appear in expensive enterprise drives for servers. However, there are plenty of precedents for new hard drive technology to gradually make its way from servers to desktops. In other words, if you need a few hundred terabytes of storage in a single drive, you've got a little patience left. The question is, when will that be?