Intel and Micron Announces Industry First QLC 3D NAND

Discussion in 'Frontpage news' started by Hilbert Hagedoorn, May 22, 2018.

  1. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    48,324
    Likes Received:
    18,405
    GPU:
    AMD | NVIDIA
  2. ObscureangelPT

    ObscureangelPT Guest

    Messages:
    552
    Likes Received:
    66
    GPU:
    Zotac GTX 1650 Supe
    The biggest question of this qlc ssds are how many writes they can handle until they go under
     
  3. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    48,324
    Likes Received:
    18,405
    GPU:
    AMD | NVIDIA
    TLC writes 3-bits per cell, with a claimed 1000 P/E cycles. QLC will write 4-bits per cell, and the interesting fact here, it'll do the same 1000 P/E cycles according to Toshiba at the time. The theory is that the NAND cell is a little bit bigger to compensate on that. Add to that improved wearing techniques and such. Time will tell though, but everybody HATED the move from SLC to MLC, then shortly thereafter TLC was the black sheep, which now has become a very proper and trustworthy NAND series. I guess It'll be the same with QLC in the beginning.

    [​IMG]
     
    Silva likes this.
  4. Kaarme

    Kaarme Ancient Guru

    Messages:
    3,511
    Likes Received:
    2,353
    GPU:
    Nvidia 4070 FE
    If you consider the nomenclature, it's quite interesting that two bits per cell is already called multi-level cell, MLC. It's as if back when it was invented, the industry thought: "That's it". So, it was just the original SLC and then the MLC, single and multi. If they had already foreseen in the more distant future there would be triple bits per cell followed by quad bits per cell, the name for the two bits per cell wouldn't have been multi, it woud be have been double, dual, or something like it. What I am saying is that if the industry itself, the one making money out of this, thought two is already the ultimate form, then surely nobody can be blamed for being suspicious when the cell is divided to more and more levels.
     

  5. Silva

    Silva Ancient Guru

    Messages:
    2,048
    Likes Received:
    1,196
    GPU:
    Asus Dual RX580 O4G
    When SLC was a norm I too hated the move on MLC because it meant you'd lose life on the SSD. But because SLC was so expensive, I would never be able to buy one.
    In the end, I was grateful for MLC and bought an SSD born from this tech. I'm really happy with it and averaging 12Gb per day it will take well over a decade for it to die on me from wear.
    That said, with the memory prices going up and people wanting more density (to compete with HDD), they had to find a way to create more density. What better way to repeat what they already knew?
    It was great people hated on the tech because that way they made sure the reliability stayed the same (bigger cells) and I'm sure we will see more in the future: 5-bit per cell, 6-bit per cell, etc.
    And it will be fine, as long as they keep the 1000 or more P/E cycles.
     
  6. Venix

    Venix Ancient Guru

    Messages:
    3,428
    Likes Received:
    1,939
    GPU:
    Rtx 4070 super
    when you talk about 2++ tb drives .... every P/E cycle should last longer no ? especially if you make this a steam library drive ! so ... in a way if they make a QLC 2tb for 100-150 euros and rate it even on 300 P/E i would personally buy one to make it my steam drive !

    also to that note so far everyone that did test on ssd's life every time the tested drives way passed their p/e cycles before they die out no ? i think the ratting they give is the worst case scenario / luck
     

Share This Page