As shown in Figure 1, data is stored in circular tracks on a disk. Within each track, a stream of data bits is recorded as regions of opposite magnetization. Each track consists of equally spaced bit cells, with a digital '1' being indicated by a boundary (called a magnetic transition) between regions of opposite magnetization within a bit cell, and a '0' being indicated by a continuous region without such a boundary. At high magnification, it becomes apparent that within each bit cell there are many tiny magnetic grains. These grains are randomly created during the deposition of the magnetic film. Each grain behaves like an independent magnet whose magnetization can be flipped by the write head during the data writing process.

Note that the boundaries between regions of opposite magnetization must occur along the boundaries between the tiny grains -- it is not possible to flip "half a grain." The fact that these magnetic transitions must follow the grain boundaries means that they are not straight, but rather meandering boundaries that merely approximate the ideal straight boundary that the write head intended to create.

If the grains are small enough, the magnetic transitions are straight enough so that it is easy to detect which bit cells contain a boundary and which do not. However, if the density is increased (which shrinks the bit cells) without shrinking the grains, the magnetic transitions become proportionally noisier, eventually preventing the readback system from accurately recovering the data. To keep the noise associated with grain boundaries small enough for reliable data detection, roughly 50~100 grains are needed per bit cell.

So the solution is obvious: Make smaller grains to support higher density recording. This is what has been successfully done for decades to allow increasing density. Today, however, grain sizes have gotten so small that further shrinkage would cause the magnetization of the individual grains to be unstable. The superparamagnetic effect tells us that when product of the grain volume V and its anisotropy energy ku fall below a certain value, the magnetization of that grain can flip spontaneously. If a significant fraction of the grains on the disk flip spontaneously, the data stored on the disk erases itself!

Since we need to keep reducing size (and therefore V) to record at higher densities, one way to maintain thermal stability would be to increase ku. However, raising ku too high results in a media coercivity which is too high -- in other words, the data would be thermally stable, but no write head would be able to generate a strong enough field to write the data in the first place.

Figure 1 (Conventional Media)

| Back to Top |