Sandisk Puts Petabyte SSDs On the Roadmap

SanDisk aims to produce petabyte-scale SSDs through its new UltraQLC platform, though the company has not specified a release timeline. The technology, it said, combines SanDisk's BICS 8 QLC 3D NAND with a proprietary 64-channel controller featuring hardware accelerators that offload storage functions from firmware to reduce latency and improve reliability.

The initial UltraQLC drives will use 2Tb NAND chips to reach 128TB capacities, with future iterations targeting 256TB, 512TB, and eventually 1PB as higher-density NAND becomes available. The controller dynamically adjusts power based on workload and employs an advanced bus multiplexer to handle increased data loads from high-density QLC stacks, the company said.

Re:

[Errol backfiring]

Well, in a way, you are right. I often use SD cards and USB drives just to carry a few files to a print shop, a 3D printer, an embroidery machine, etc. But these things get larger and more expensive, while it just does not make any sense in the use I have for them.

Off course we are talking about hard drives here, but at some point you want specialized hardware if you really need that amount of storage. There are special RAID configurations for video, special data storage devices for large data sets of resea

Re:

[commodore73]

Yeah, not my best joke, and now it's stuck there for eternity.

Re:

[OrangAsm]

640k is enough for anyone.

I agree, 640k GB is pretty good.

Blech. QLC.

[dgatwood]

Storing a petabyte of data on flash devices with maybe a few hundred to a thousand write cycles before they fail doesn't sound like a very bright idea to me. In fact, it sounds like a great way to lose a s**tload of data very quickly.

Let me know when 3D flash memory actually hits the market in sufficient quantities for the entire industry to go back to MLC or SLC so that drive reliability won't suck anymore.

Re:Blech. QLC.

[ihadafivedigituid]

You won't even write it once during your lifetime.

Even at 10GB/sec, those 500 writes would take nearly 580 days. At 1GB/sec, it's nearly 16 years of nonstop writing.

Re:

[TheMiddleRoad]

How about at 2400 baud? Asking for a friend.

Re:

[ihadafivedigituid]

Did Comcast upgrade their network?

Re:

[Hoi Polloi]

And unless you are talking about massive files (which wouldn't be a good idea for a db or even fit in memory) you'd have a huge file organizational issue.

Re:

[kwelch007]

And unless you are talking about massive files (which wouldn't be a good idea for a db or even fit in memory) you'd have a huge file organizational issue.

I suppose that's true, but that has nothing to do with the underlying storage hardware. There are already proven solutions to that problem.

Re:

[MooseTick]

So it could be great for storing videos? Perhaps stuff like security camera footage?

Re:

[thegarbz]

It would sound like a bad idea to you until you consider that modern wear levelling algorithms make such an SSD insanely long life. The thing about a Petabyte of data is that you either want to store it stably for a long period, have massive redundancy due to its constant changing mission critical nature (at which point you want a lightning fast device to recover from), or you have a temporary use for the data such as in simulation where the speed is again critical but the data is not.

An SSD is a perfect de

Re:

[AmiMoJo]

Just reduce the usable capacity and have plenty of spare blocks.

I'm more interested in Huawei's new flash/tape hybrid. Massive storage with a built in large flash cache. The tape market needs some disruption, it's too expensive, and this looks like it will be very practical and cost effective.

Re:

[karmawarrior]

Wouldn't it have the same failure mode that virtually all other flash devices have, eg firmware detects when it's getting close to the write limit and switches to read only when that happens?

That said, does QLC have the "Must be powered on regularly or loses data" issue that other flash technologies have?

Re:

[pandaman9000]

"Wouldn't it have the same failure mode that virtually all other flash devices have, eg firmware detects when it's getting close to the write limit and switches to read only when that happens?"

It seems logical for the "read only when full" part to be accurate. Perhaps it could actually be when full and wear levelling room is X% exhausted....
I'm not aware of whether all current SSD drives have this capability, but my Samsung 990 Pro drives can have space defined as unusable for data other than that moved fro

Re:

[karmawarrior]

On the unpowered flash issue:

It's been a while since I looked it up and the Internet is not being that helpful this morning, with one site suggesting consumer grade SSDs last between 1 and 5 years without power, and another implying regular SSDs store data for "5-10 years stored at a normal temperature". That said, I know the ratings for SD cards are usually lower, to the point it's a popular warning to photogs etc.

Those are practical limits, in terms of spec'd limits apparently JEDEC218 specifies a one yea

Re:Blech. QLC.

[paradigm82]

The JEDEC requirement is 1 year for an unpowered consumer drive as you say. However, the period the drive is able to retain data decreases with the amount of write cycles it has sustained. The JEDEC requirement has to be satisfied after the disk has sustained the rated number of writes, so the retention period is likely higher than 1 year for a fresh drive. Having said that, wear mechanisms are complicated and temperature plays a major role: The hotter it is when you write the data, the longer it will last. On the other hand, the hotter it is AFTER you wrote the data, the sooner it will perish. At any rate, the 1 y guarantee depends on the drives ability to apply error correction algorithms, the first bits are lost much sooner. There's also differences in how drives handle it when they cannot fully recover the data - do they return an error, or do they just return the best guess and increase some counters? Another complication: It is not enough to just power on the drive for a second... the 'powered' part depends on the drive running background routines that scan for aged areas and refreshes them. The algorithms for this are not documented. Since reading itself wears the data (see "read disturb" effect) and at any rate spends power (especially relevant in laptops) and also takes time for a full drive scan, it is hard to know when you can be sure. Personally, I have the drives in my main machine so they are often powered. But in addition, I backup, completely reformat (SATA security erase or the NVMe equivalent) the drive every 2-3 years and copy all the data back. I still have an almost 12 years old 840 EVO chugging along without problems. It was the first SSD to highlight the problem with retention due to it being a TLC architecture but without V-NAND, meaning data retention is especially low... in fact slowdowns in accessing old data can be seen in less than a year, due to the error correction algorithms working hard to interpret the data.

Re:

[juancn]

At that size, you could do append only and keep literally everything. Never overwrite (except to permanently erase).

It may make sense to have new filesystem.

SSD size isn't an issue

[JamesTRexx]

I'd rather see a faster bus speed between RAM and CPU. After all, that's still the bottleneck for handling large amounts of data.

Until your girlfriend throws it away

[zawarski]

Heâ(TM)ll of a time finding it in a landfill.

What size?

[algaeman]

128TB with 2Tb chips is 512 chips. Will that even fit in a 3.5" disk form factor?