Stories
Slash Boxes
Comments

SoylentNews is people

Micron’s New 3D NAND Flash Could Usher in a Rapid New Generation of SSDs

posted by janrinok on Tuesday May 17 2022, @10:12AM   Printer-friendly
Hardware

upstart writes:

Micron's new 3D NAND flash could usher in a rapid new generation of SSDs:

Micron has revealed it has developed 3D NAND flash with a whopping 232 layers, which will enter full-scale production later this year.

[...] The company has not yet provided performance specs for its 232-layer device, but implied that speeds will exceed those of its current 3D NAND products, paving the way for rapid and capacious new SSDs.

NAND flash is a type of non-volatile memory that features in all kinds of storage devices, from memory cards, USB sticks and portable drives to SSDs for devices and servers.

The general idea behind NAND flash development is to reduce cost per capacity and increase storage density, effectively eliminating the use cases for traditional hard disk drives.

[...] SSDs powered by the new 3D NAND flash are expected to come to market at some point in 2023.

Original Submission

This discussion has been archived. No new comments can be posted.
Micron’s New 3D NAND Flash Could Usher in a Rapid New Generation of SSDs | Log In/Create an Account | Top | 8 comments | Search Discussion
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)

  • (Score: 1, Interesting) by Anonymous Coward on Tuesday May 17 2022, @11:03AM (4 children)

    by Anonymous Coward on Tuesday May 17 2022, @11:03AM (#1245605)

    I'm quite sure all those "whopping 232 layers" will be whopping dog slow in QLC or TLC mode, irrespective of all this empty advertising.

    • (Score: 4, Informative) by takyon on Tuesday May 17 2022, @04:04PM (3 children)

      by takyon (881) <takyonNO@SPAMsoylentnews.org> on Tuesday May 17 2022, @04:04PM (#1245689) Journal

      Most of the fast PCIe 5.0 SSDs are going to use at least TLC NAND. I'm pretty sure the vertical nature of 3D NAND improves parallelism, along with newer memory controllers with more channels.

      Put SLC and/or DRAM cache in the drive, and write speeds can be sustained for typical consumer workloads.

      The real problem facing these SSDs will be heat. There's going to be active cooling on SSDs [pcworld.com].

      --
      [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]

      • (Score: 0) by Anonymous Coward on Tuesday May 17 2022, @05:49PM (2 children)

        by Anonymous Coward on Tuesday May 17 2022, @05:49PM (#1245754)

        for typical consumer workloads.

        Have movies in 4k officially become "atypical" right this morning, or what? Have backups become that thing too?
        All that "cache" fakery goes to shit (exposes the shit behind it, that is) anytime a copying of a few GBs happens. All that for the laughable 50% space increase of TLC above MLC?

        Why not give user an option to convert the entire drive to "cache" mode, for a corresponding reduction in available space? The user paid for that silicon already, WHY force on them crappy speed AND crappy durability in exchange for paltry extra space they might not ever need at all?

        • (Score: 2) by takyon on Tuesday May 17 2022, @07:04PM (1 child)

          by takyon (881) <takyonNO@SPAMsoylentnews.org> on Tuesday May 17 2022, @07:04PM (#1245776) Journal

          They can already do that.

          https://www.anandtech.com/show/13421/the-mushkin-source-sata-ssd-review/5 [anandtech.com]

          The Mushkin Source [500 GB] lasts a very long time before its SLC cache is filled. While the Toshiba TR200's write speed ends up in the gutter very quickly, it will be almost impossible for a real-world consumer workload to fill the cache of the Source unless the drive starts out nearly full. From an empty drive, the apparent initial SLC cache size is well over 150GB—essentially the entire drive operating as SLC until free space runs out.

          https://www.anandtech.com/show/15887/the-samsung-870-qvo-1tb-4tb-ssd-review-qlc-refreshed/2 [anandtech.com]

          The SLC caches on the 870 QVOs run out right on schedule, at 42 GB [1 TB] and 78 GB [4 TB]. Write performance drops precipitously but is stable thereafter, for the rest of the drive fill process. This behavior hasn't changed meaningfully from the 860 QVO.

          https://www.anandtech.com/show/16012/the-sk-hynix-gold-p31-ssd-review/2 [anandtech.com]

          The SLC write cache in the 1TB SK hynix Gold P31 runs out after just over 100GB of writes. After the SLC cache fills up, the Gold P31's sequential write performance becomes highly variable, ranging from about 1.4 to 2.3 GB/s with little change in character across the entire TLC filling phase. There are no obvious patterns of periodic garbage collection cycles visible at this scale.

          --
          [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]

          • (Score: 1, Insightful) by Anonymous Coward on Tuesday May 17 2022, @08:40PM

            by Anonymous Coward on Tuesday May 17 2022, @08:40PM (#1245798)

            Thanks for the first link. That drive where "essentially the entire drive operating as SLC until free space runs out" is maybe good enough if not a lie: make a cache-sized partition, leave the rest unallocated, use as SLC. Still, an option to configure the drive into SLC mode, making its reported size cache-sized and preventing any "garbage collection" from SLC into TLC storage mode, would be nice to have, and cost nothing to the manufacturer.

  • (Score: 0) by Anonymous Coward on Tuesday May 17 2022, @01:33PM (1 child)

    by Anonymous Coward on Tuesday May 17 2022, @01:33PM (#1245626)

    If disk access speeds start approaching RAM access speeds then there is no need to move stuff from disk to RAM, your disk IS the RAM.

  • (Score: 5, Insightful) by looorg on Tuesday May 17 2022, @01:35PM

    by looorg (578) on Tuesday May 17 2022, @01:35PM (#1245628)

    That is nice. But for speed I think I'm fine. I want them to live longer and not just magically die in a *poof* anymore. Fix that part instead.

(1)