HP Researchers Develop The Memristor Which Could Revolutionize Electronics

Posted on 06 May 2008

Memristor

Theoretical physics are one of the most important discoveries that the humanity ever done because we rely on them to develop technologies that the medieval humans thought of them as witch-crafts. Hewlett Packard has proudly informed the whole world that they have developed a revolutionary electrical component.

This new nanoscale component is called a memristor, but I have to say that it’s not entirely a new-comer as an electrical engineer named Leon Chua first explained it in theory in the early 1970s. This memristor is very functional and really interesting because it can work in two ways, the first, the digital mode which is simpler because memory cell is either “on” or “off”. The other way is analog mode in which the memory cell can have values between the “on” and “off”. The values of the analog mode are determined by the number of times in which the cells receive electrical signals that make these values to grow (this process is very similar to the way that neurons are building stronger memories when they are stimulated).

According to the HP researchers, the memristor does not require power to retain data or information, and it can store info denser than a hard-drive and although for the moment it cannot access it as fast as the RAM memory of a computer, the developers are pretty sure that in the future it can replace both HDD and RAM.

Going back to the fact that Leon Chua discovered it almost 40 years ago, but the experimenters did not found it - this is explained very well by the fact that scientists first suggest a particle and the researchers might find it or not and the most recent example is the quest for the theoretical Higgs boson known as the “God particle”. The HP researchers were very lucky to “catch” the memristor in action while they were studying the electrical properties of the nanoscale materials when they noticed something that was behaving like a memristor - they made some adjustments and Chua’s dream theory was true.

Now, where this discovery can lead to? Well, according to Stan Williams, the leader of the HP research team, the memristor will be available for PCs in the next few years as the cache which lies between the hard-drive and the DRAM. How this works? The hard-drive will load the key data into memristor’s cache, then dump it into the DRAM faster than moving it directly from the hard-drive which means that we will be able to boot and open large files and folders with the speed of light. This is only for the next few years, says Williams, as his masterplan is to substitute the hard-drive and the RAM of the PC.

Another important part of this discovery can lead to a CPU with many processing cores very different of a digital computer. The logic of a digital computer can be either yes or no which makes it very difficult to recognize a face, but with Williams working on an electronic mind which will make a PC to think exactly as a human, facial recognition will be a piece of cake. Williams does not think this is possible for the moment but “we’re going to build a brain anytime in the next decade” and even if they will not succeed, they are not disheartened because lets remember the fact that memristor took almost 40 years to finally be discovered.

Leon Chua named this particle memristor from memory resistor and back then, he said that the electric resistance will grow its levels when stimulated and it will remember it until another stimulation and the bad thing was that nobody ever saw it.

The memristors developed by HP measure about 15 nanometers across, enough to store information as much as a hard-drive (about 100GB per square centimeter). If it were to believe HP researchers, this memristor could be even smaller - between 4 and 2 nanometers.

I am really excited by this discovery because a just a few days ago I was totally convinced that those android movies are pure fantasy, but now I see that I was totally wrong. Everybody will love the next generation of computers and I am not sure about the computer thinking like a human.

This post was written by:

Dragos Pirvu - who has written 71 posts on DoSci - Science Blog.


Contact the author

1 Comments For This Post

  1. JTankers says:

    Very cool technology. Hi Electrical Engineers, I have a challenge for you… (I am a computer programmer and cosmology is my hobby…)

    LHCConcerns.com will pay $500.00 US to the best proposal that can reasonably prove 5% or less Risk of Planetary destruction from Micro Black Holes.

    The contest will conclude in a vote by site visitors on all reasonable proofs received, all proofs will be published and the contest will end not sooner than May 20th. (LHCConcerns will make the final call on best proposal that reasonably proves 5% or lower risk from micro black holes being created by the Large Hadron Collider).

    You may prove that ANY ONE of the following or provide any other reasonable Proof or method to prevent Micro Black Holes from being created by the Large Hadron Collider or prove that they are harmless!

    1. The Large Hadron Collider will not make micro black holes.
    2. Micro black holes created will be sent safely into space.
    3. Micro black holes will evaporate.
    4. Micro black holes will take more than 2 billion years to accrete the Earth. (If you can only prove a lesser time frame, then the prize will be reduced proportionately…)
    5. Any form of cosmic ray argument that proves 5% risk or lower.
    6. Find a way to make the Large Hadron Collider safe from creating micro black holes (we already requested different speed collissions or different mass collisions, LSAG told us it was not possible, they already thought of it).

    It is harder than it looks, the LHC Safety Assessment Group (LSAG) could not produce a safety report… (CERN and LSAG are still using the 1999 RHIC safety report that does not even address what might happen if micro black holes were created, because they did not know that it was possible at that time. We are also being generous on the 2 billion years, we want to be reasonable)

    JTankers
    LHCConcerns.com

Leave a Reply