Wednesday, November 12, 2008

University embarks on carbon nanotube data storage project, promises DRAM-like non-volatile memory

By Rick C. Hodgin
Wednesday, November 12, 2008 03:30

Nottingham (UK) - Researchers at The University of Nottingham, one of UK's Top 10 universities, also ranking in the world's Top 100, stated yesterday that Project Nanodevice is underway. Their goal is to create molecular memory built of telescoping carbon nanotubes. "In this project a new device for storing information will be developed, made entirely of carbon nanotubes and combining the speed and price of dynamic memory with the non-volatility of flash memory."

Telescoping a carbon nanotube

The idea sounds simple enough; two carbon nanotubes of slightly different size, one resting inside the other like a two-member set of matryoshka (Russian dolls where each one has a smaller one inside). Electrical current will pass through the outer tube forcing the inner tube to telescope in or out. When out it will make contact with a remote electrode, thereby completing a circuit to create a binary one. When retracted the circuit will be broken - a binary zero.

An artist's rendition of carbon nanotube memory in theory. (a) shows a full extended telescope and completed circuit representing a binary one. (b) shows a retracted telescope and an incomplete circuit creating a binary zero. Single-atom thick walls allow massive storage potential on the order of 10-100x more dense than modern flash memory with read/write speeds rivaling DRAM.

This kind of memory will be a physical displacement of matter, meaning something has to move in order to switch states. Tiny, rolled sheets of graphene make up the carbon nanotubes, and these exhibit molecular properties which make the movements extremely fast and reliable, at least theoretically.

Non-volatile, fast and friendly

Another advantage of this system is that the memory will be non-volatile. Just like flash memory today, it won't need any power to maintain its state. It should also be extremely resistant to G-force induced state changes, such as accidental droppage.

Even early generations of this technology should be as fast or faster than modern DRAM. Future computers using this kind of memory won't have a separate memory and hard disk for storage. Theirs will be a unified memory architecture built around this kind of storage medium, a new design paradigm for the instant on computer, one capable of continuous processing and data storage without ever swapping memory out to hard disk through paging. This one fact alone would greatly speed up our computer experience today.


The project is being led by Dr. Elena Bichoutskaia, who said, "The electronics industry is searching for a replacement of silicon-based technologies for data storage and computer memory. Existing technologies, such as magnetic hard discs, cannot be used reliably at the sub-micrometre scale and will soon reach their fundamental physical limitations."

Her goals, and the goals of the research teams working on this project, is a new memory device. According to Bichoutskaia, a new carbon nanotube memory product will be produced, one that will replacing DRAM and flash. With research of this nature there are no timeframes.

Personally, I suggest a name for this creation, one in keeping with the finest traditions of existing memory naming conventions: CRAM (Carbon-nanotube RAM). "How many smaller nanotubes can I cram inside the bigger ones?" Perhaps future generations could move away from binary computers into ternary (or beyond) by having multiple tubes, like a real telescope.