At CEATEC Japan, TDK was demonstrating HAMR (heat-assisted magnetic recording), a technology designed to dramatically increase the data density of mechanical hard drives. Hard drives are typically limited by how much bits of information they can safely store per square inch of each platter surface (HDDs can feature many platters or disks). Since the information is saved as a magnetic polarity, attempting to pack too much information in a tight space may result in weird analog interferences that ends up breaking down individual bits of information and effectively corrupt the data.
TDK has found that magnetic surfaces used in hard drives react differently at higher temperatures, and that by heating the disk surface, they were able to safely pack more information per square inch. To do that, they use a laser that heats up the surface ahead of the magnetic write/read head so that the surface comes at the ideal writing temperature when the HDD head is right on top of it. The challenge is to find the right balance of heating/cooling speed, without inducing any damage to the platter. That’s exactly what TDK claims to have done.
Once the surface has cooled, it works just like any other HDD, and we’re back to what we know very well. Since the density has increased, chances are that peak throughput will also increase. Additionally, TDK points out that its HAMR technology will also save power on a “per GB” basis since less drives are used for a given storage capacity.
In conclusion, TDK’s HAMR pushes the limit of platter-based magnetic storage disks, but doesn’t lift the “wall” that this technology is supposed to hit at some point. To be fair, although it’s common wisdom that SSDs are the future, semiconductors are also presumed to be going to hit a wall when their manufacturing process gets closer to a few atoms wide per electric wires. In the meantime, engineers in both industries are working hard to push those limits further and HAMR is a great example of how seemingly small changes can turn into a huge deal.