Slashdot Mirror


Using DNA To Build DRAMs

SailorBob writes "A research team at the University of Minnesota has demonstrated a technique using DNA crystals to build 20 angstrom 1 bit single electron memory cells. The researchers claim that this technique could be used to build memory with a 10 trillion bits/cm2 density, which would mean a 6.4 Terabit DRAM."

2 of 21 comments (clear)

  1. not exactly DRAM by Matthias+Wiesmann · · Score: 4, Interesting
    They article only compares the DNA system with DRAM for memory density. They don't talk about building DRAM using this technology.
    • The D in DRAM stands for dynmic, this is memory that needs to be update at regular intervals. I'm not sure this DNA technology will require refreshes (this is good).
    • In the article, they mention that their current plans are not for random access (those are the R and A in RAM.) I think this technology is closer to CCD memory where data can be fed and read in a sequential manner. For certain applications (like storing video data), this is sufficient, but for computer main memory, this is not well suited.
    • So basically, they will do DRAM, but without the DRA part. This is called memory.
    1. Re:not exactly DRAM by Simon+Field · · Score: 3, Interesting


      My thought when I first read this was not memories, but high efficiency (>85%) solar cells.

      The folks working on optical rectennas are to the point where their remaining problems are with making metal-insulator-metal tunneling diodes that are small enough and fast enough.

      They are close to demonstrating a working device, and it looks to me like DNA scaffolding could not only make it easier to build the MIM diodes, but might make fabrication of the entire device easier and faster. Instead of ion-beam implantation, you grow the 2D crystals in a vat.

      The idea is to make an array of tiny dipole antennas, with a diode at the feedpoint. When light hits the antennas, an AC current is generated, then rectified by the diodes.

      Making the antennas has been possible for some time now (500 nanometers). The work on making the diodes is nearly ready. Now it looks like we have a possible cheap and efficient means of constructing them on a large scale.