Millipede is another new technology that promised much but so far has failed to deliver. As of 2018, because the cost and density of flash storage has improved so much, no commercial millipede product has been made available so far
Millipede is a little like a hard drive, in that it moves storage under fixed heads, but the difference is that millipede uses a large array of fixed heads in a grid pattern, so it can write and read lots of data in parallel. The data storage bits are microscopically small, which means it can store data at a very high density. OK, so why is it not out there yet?
Millipede is based on the nanometer sharp tips used in an atomic force microscope (AFM). Data bits are stored as a pit, or the absence of one, in the surface of a thermo-active polymer, or 'data sled' as it is called.
The main issue with millipede is that it works with heat, the nanoprobes must be heated to 300 degrees centigrade to read data, and 400 to write data. In short, data is written by heating the relevant probe tips and pushing them into the data sled. This softens the surface and leaves a small depression. Data is then read by pushing the probe heads into the data sled at a lower temperature. Where 'holes' exist, the hot probe enters the data sled and heat leaks away, so lowering the probe resistance. So a low resistance, or 'hole' is then a data '1' and a high resistance, or 'no hole' is a data '0'. This heat process requires quite a bit of power, and is relatively slow, but the advantage of millipede is that lots of reads or writes can happen in parallel so the overall process is speeded up.
The problem for Millipede is that disk and solid state technology is still improving rapidly, so it is hard to find a point where millipede can enter the market at a suitable economic level. In theory, a Millipede can store data at 1Tb per square inch and in practice, working models store data at 800Gb per square inch, which is more magnetic disks, but not significantly more.
IBM foresees two main applications for millipede; applications that need very small storage devices, like watches or mobile phones, and applications that require terabytes of storage. Millipede could scale up either by producing chips with say 1,000,000 cantilevers, but it would be easier to use lots of smaller millipede chips in parallel. Another approach would be to create the polymer coat on a modified hard disk, which could result in a 2.5 inch disk with several terabytes capacity.
However just as you think you can write a technology off, a new variant appears. Scientists at Pohang University of Science and Technology in Kyungbuk, Korea, have developed a new polymer called a 'baroplastic' which works on pressure rather than heat. The normally hard baroplastic polymer softens under pressure and so instead of heating up a needle tip, you just push it into the data sled to create a data store. In this system, pressure replaces heat in the reading and writing of the millipede storage chip, thus allowing the process to occur at room temperature. Then again, this technology has its own problems as the constant pushing into the data sled wears the needle heads. The scientists are now looking at combining the baroplastic polymer with some other substance that does not wear the heads.
Another technology that promises fantastic memory densities, up to 500 Terabits per square inch, is chlorine atoms on a copper surface. This is under development at the Technical University of Delft. A single chlorine atom can occupy one of two positions on a surface of copper atoms, and its position will represent a single bit, a '1' or a '0'. The researchers used a scanning tunneling microscope (STM), which pushes the atoms on the copper surface one by one into the desired position. The researchers say 'Every bit consists of two positions on a surface of copper atoms, and one chlorine atom that we can slide back and forth between these two positions. If the chlorine atom is in the top position, there is a hole beneath it - we call this a 1. If the hole is in the top position and the chlorine atom is therefore on the bottom, then the bit is a 0.'
However we should not get too excited about this technology yet. The prototype was just held 1 KB, and the access time was measured in minutes. Also, the chip needed to be kept at a temperature of -196C.