1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Taiwan engineers defeat limits of flash memory

Discussion in 'Science & Technology' started by Drone, Dec 2, 2012.

  1. Drone

    Drone

    Joined:
    Sep 1, 2010
    Messages:
    2,571 (1.77/day)
    Thanks Received:
    1,274
    We heard of HAMR and salty HDDs and now it's getting even better, this time for NAND. Flash memory can get its life prolonged :D Significantly ...

    Self-healing lol just like the The Ol' Canucklehead.

    And guess what ... life-giving workaround is ... heat

    Source
  2. de.das.dude

    de.das.dude Pro Indian Modder

    Joined:
    Jun 13, 2010
    Messages:
    7,561 (4.94/day)
    Thanks Received:
    1,966
    so.. a normal SSD chip dies after 10000 cycles?
  3. McSteel

    McSteel

    Joined:
    Nov 19, 2012
    Messages:
    572 (0.90/day)
    Thanks Received:
    275
    @de.das.dude: A chip is a collection of flash cells. Cells are organized into pages, typically ~1000 cells per page. Pages are then organized into blocks, 32, 64, 128 pages per block exist. Chips can have from one to eight blocks.

    But to answer your question, a mediocre SSD will endure about 10GB of random writes each day for 3 years before starting to wear out. Even then, as writes become impossible, reads will still be possible for some time, meaning you can copy your data before the drive fails completely.

    @Drone: I don't see how flash-heating the flash (heh) will help with anything, long-term... All heat does is randomize the electron distribution between the substrate and the floating gate. Flash cells degrade because electrons that are tunneled into the floating gate from the substrate can leave the floating gate at random (on average, that's literally a one-in-a-billion chance when accessing a cell). Also the substrate "evaporates" slowly, as the electrons rushing through the depletion region can sometimes push out substrate atoms, thus damaging/degrading it. All other degradation occurs as a direct consequence of the way NAND flash is organized and engineered. Until I see where exactly the heat is applied, how, and why it should help with anything, all this is simply baseless sensationalism.
  4. Drone

    Drone

    Joined:
    Sep 1, 2010
    Messages:
    2,571 (1.77/day)
    Thanks Received:
    1,274
    Baseless what? I don't get yr point :confused: They'll propose and demonstrate their "high-temperature" flash at presentation.
  5. McSteel

    McSteel

    Joined:
    Nov 19, 2012
    Messages:
    572 (0.90/day)
    Thanks Received:
    275
    Meaning they're still working on it. And until they provide a very detailed explanation as to why heat even works as a solution (or even a workaround), and until there is a working prototype, I call shenanigans.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)

Share This Page