Toshiba last week announced its first 3D NAND flash memory chips featuring QLC (quadruple level cell) BiCS architecture. The new components feature 64 layers and developers of SSDs and SSD controller have already received samples of the devices, which Toshiba plans to use for various types of storage solutions.

Toshiba’s first 3D QLC NAND chips feature 768 Gb (96 GB) capacity and uses 64 layers, just like the company’s BICS3 chips with 256 Gb and 512 Gb capacities launched in 2016 and 2017. Toshiba does not share further details about its 3D QLC NAND IC (integrated circuit), such as page size, the number of planes as well as interface data transfer rate, but expect the latter to be high enough to build competitive SSDs in late 2018 to early 2019 (that’s our assumption). Speaking of applications that Toshiba expects to use its 3D QLC NAND ICs, the maker of flash memory mentions enterprise and consumer SSDs, tablets and memory cards.

Endurance++

Besides intention to produce 768 Gb 3D QLC NAND flash for the aforementioned devices, the most interesting part of Toshiba’s announcement is endurance specification for the upcoming components. According to the company, its 3D QLC NAND is targeted for ~1000 program/erase cycles, which is close to TLC NAND flash. This is considerably higher than the amount of P/E cycles (100 – 150) expected for QLC by the industry over the years. At first thought, it comes across a typo - didn't they mean 100?. But the email we received was quite clear:

- What’s the number of P/E cycles supported by Toshiba’s QLC NAND?
- QLC P/E is targeted for 1K cycles.

It is unclear how Toshiba managed to increase the endurance of its 3D QLC NAND by an order of magnitude versus initially predicted. What we do know is that signal processing is more challenging with QLC than it is with TLC, as each cell needs to accurately determine sixteen different voltage profiles (up from 2 in SLC, 4 in MLC, and 8 in TLC). 

The easiest way to handle this would be to increase the cell size: by having more electrons per logic level, it is easier to maintain the data and also read from it / write to it. However, the industry is also in a density race, where bits per mm^2 is an issue. Also, to deal with read errors from QLC memory, controllers with very advanced ECC capabilities have to be used for QLC-based SSDs. Toshiba has its own QSBC (Quadruple Swing-By Codes) error correction technique, which it claims to be superior to LDPC (low-density parity-check) that is widely used today for TLC-powered drives. However, there are many LDPC implementations and it is unknown which of them Toshiba used for comparison against its QSBC. Moreover, there are more ECC methods that are often discussed at various industrial events (such as FMS), so Toshiba could be using any or none of them. The only thing that the company tells about its ECC now is that it is stronger than 120 bits/1 KB used today for TLC. In any case, if Toshiba’s statement about 1000 P/E cycles for QLC is correct, it means that that the company knows how to solve both endurance and signal processing challenges.

The main advantage of QLC NAND is increased storage density when compared to TLC and MLC, assuming the same die size. As was perhaps expected, die size numbers were not provided. However, last year Toshiba and Facebook talked about a case study QLC-powered SSD with 100 TB of capacity for WORM (write once read many) applications and it looks like large-capacity custom drives and memory cards will be the first to use QLC for cold storage. P/E cycles and re-write endurance isn't a concern for WORM at this stage.

Toshiba has begun to sample its 3D QLC NAND memory devices earlier this month to various parties to enable development of SSDs and SSD controllers. Taking into account development and qualification time, Toshiba plans to mass produce its BiCS3 768 Gb 3D QLC NAND chips around the same time it starts to make its the next generation BiCS4 ICs. The latter is set to hit mass production in 2018, but the exact timeframe is yet to be determined.

Related Reading:

Source: Toshiba

Comments Locked

36 Comments

View All Comments

  • Freakie - Monday, July 3, 2017 - link

    Read hot, store cold is the rule for SSDs. The hotter the environment they are in while they are active, the less wear and tear on the cells because it's easier to move the electrons around. The colder you store them, the harder it is for electrons to leak.
  • Freakie - Monday, July 3, 2017 - link

    I guess that should be "Use hot, store cold" not read hot. But you get the point.
  • beginner99 - Wednesday, July 5, 2017 - link

    Why would you use an SSD to archive and not a cheapo hdd at a fraction of the cost and with less issue of losing data?
  • HomeworldFound - Wednesday, July 5, 2017 - link

    I generally wanted know how much of a bad idea that was, ignoring the cost, and how far away the industry is from having non-mechanical drives for storage purposes where some of us don't like mechanical drives very much.
  • HomeworldFound - Wednesday, July 5, 2017 - link

    Oh, and one thing I meant to mention were the transfer speeds and noise etc. After all of these years I'm very tired of the time spent, the noise and vibration.
  • azazel1024 - Thursday, July 6, 2017 - link

    The answer is, yes. The spec for P/E cycles is 1yr of storage persistence once that is reached. So if it is a 5k P/E drive, at 5k P/E cycles, the storage should persist for a year, before it wouldn't be retrievable. Of course that is only a very good chance of the data being recoverable without errors, not a perfect guarantee.

    The lower the number of P/E cycles on the drive, the longer the data can be retrained error free. It isn't a log function and now I am too lazy to go pull up the white paper I had been reading on it for another Anandtech article I commented on recently. But an MLC drive they were looking at had something like 1/30-1/10th the raw read error rate of P/E count of 1 vs P/E count of 5,000 (I assume based on 2x nm flash drive, that was roughly the max rated P/E count for MLC).

    So figure 10-20 years of data retention.

    IF KEPT AT ROOM TEMPERATURE. NAND flash very rapidly loses charge at higher temperatures. At 70C, about what you could expect in a hot car on a sunny day with a black SSD enclosure with the sun right on it, it loses charge something like 300x faster. Even at 50C it accelerates data loss by something around 30-40x and 40C it is about 10-15x. This is all versus 20C storage.

    So if you live in the tropics and leave SSDs laying around on a shelf in an unconditioned building, probably figure on a barely used drive you might get 1-2 years of data retention. On an abused drive, weeks. In the arctic you might get centuries of data retention.
  • MamiyaOtaru - Tuesday, July 4, 2017 - link

    won't leak faster, but it's a lot easier to tell the difference between 0 and 1 (SLC) than it is to tell the difference between 0110 and 0111 (QLC). That is to say you can leak a lot before 1 looks like 0, but don't have to leak nearly as much before 0111 looks like 0110
  • Pork@III - Tuesday, July 4, 2017 - link

    Planned obsolescence.

    ..."QLC that is..."

    ....garbage!
  • Bullwinkle J Moose - Tuesday, July 4, 2017 - link

    Planned obsolescence ???

    ..."TOSHIBA that is!..."

    Endurance claims appear higher than normal and consumer availability is conveniently after the Company will be sold to the highest bidder

    Anyone sampling these now should torture the crap out of them to see if the endurance claims are anywhere near accurate before this Company Vaporizes as fast as its claims
  • Alexvrb - Tuesday, July 4, 2017 - link

    I'm all for QLC... for secondary storage only. Basically replacing a storage HDD used for media with a QLC SSD. Current TLC drives are far too expensive to replace multi-TB HDDs. Anyway, in that scenario QLC would be good enough, you're not going to be hammering the drive with writes.

    For primary storage, a really good TLC based drive is a bare minimum. QLC, even supposed high endurance QLC, isn't good enough. Even if the endurance is truly as good as they claim, performance won't be. Especially when discussing PCIe-linked NVMe solutions (mostly M.2).

    I envision a rig with an PCIe 4.0/5.0 main drive (possibly PCM/ReRAM), and a high-density QLC drive for mass storage.

Log in

Don't have an account? Sign up now