Intel getting into the TLC SSD game??? Say it ain't so...

[Magic Carpet]

Classic race to the bottom example.

Yeah, like with the laptops. They used to be quality, now we see lots of $200-300 models made of junk. Same idea.

Hell, even iPhone SE is getting there, lmao (still quality though). Step by step.

 

[fleshconsumed]

Yeah, like with the laptops. They used to be quality, now we see lots of $200-300 models made of junk. Same idea.

Hell, even iPhone SE is getting there, lmao (still quality though). Step by step.

Well, laptops are a bit different. You can still buy a quality laptop if you're willing to pay. Dell XPS13 has aluminum shell, carbon fiber palm wrest, minimalistic screen and QHD screen. You can still have quality.

However, I'm not sure if SSDs will follow the same direction. In case of laptops it's easy to see what you're paying big money for. In case of SSDs and reliability, not so much. I'll keep buying MLC for as long as it's available. I value reliability too much to buy TLC.

 

[Stuka87]

I just bought another Intel 730, because production of them has been stopped, and they are getting harder to find. The days of enterprise grade consumer drives are over :/

 

[hojnikb]

I'm just wondering at this point when QLC will hit the scene with 5 total host write lifetimes of the cells...

qlc already exists.

 

[hojnikb]

Yeah, the recommendation for SSDs used to be, just buy the cheapest one, all of them are much faster than HDDs. This doesn't apply any more. Getting lower performance from a cheap drive doesn't bother me, but the data retention issues and performance degradation over time are huge issues that should have been ironed out at the testing/development stage long before they were released to consumers.
Maybe Intel have finally mastered 16nm TLC, but I'd rather not risk it.

unlikely, given they are using 3rd party components.

 

[cbn]

I'm just wondering at this point when QLC will hit the scene with 5 total host write lifetimes of the cells...

qlc already exists.

http://www.zdnet.com/article/toshibas-quad-level-flash-coming-next-year/#

QLC enables a single cell to hold 4 bits of data. That increases storage density by a third, without the extra costs involved with 3D manufacturing. Cheaper flash, for almost free.

But there are downsides. With each increase in bit density - from SLC to MLC to TLC - the cell becomes more sensitive to wear and provides fewer writes before it can be written no more.

In the case of QLC, Mr. Oshima suggested the limit would be about 500 writes, which, surprisingly, the Google SSD study found isn't much less than what SSDs in heavy use see.

Translation: should be fine for consumer and, with proper controller firmware, enterprise use.

What's more, Mr. Oshima predicted that Toshiba would be able to offer 2.5" 88TB flash drives that combined QLC, BiCS, and much higher z-axis chips. That's a few years out, but definitely on the Toshiba roadmap.

 

[hojnikb]

http://www.zdnet.com/article/toshibas-quad-level-flash-coming-next-year/#

sandisk did it back in 2009 already
https://www.sandisk.com/about/media...it-x4-(4-bits-per-cell)-nand-flash-technology

 

[cbn]

Some more info on the Toshiba QLC (this is an older article from August 2015):

http://www.pcper.com/news/Storage/FMS-2015-Toshiba-Announces-QLC-4-bit-MLC-3D-Archival-Flash

That's right - QLC (Quadruple Level Cell), which is also 4-bit MLC, has been mentioned by Toshiba. As you can see at the right of that slide, storing four bits in a single flash cell means there are *sixteen* very narrow voltage ranges representing the stored data. That is a very hard thing to do, and even harder to do with high performance (programming/writing would take a relatively long time as the circuitry nudges the voltages to such a precise level). This is why Toshiba pitched this flash as a low cost solution for archival purposes. You wouldn't want to use this type of flash in a device that was written constantly, since the channel materials wearing out would have a much more significant effect on endurance. Suiting this flash to be written only a few times would keep it in a 'newer' state that would be effective for solid state data archiving.

The 1x / 0.5x / 6x figures appearing in the slide are meant to compare relative endurance to Toshiba's own planar 15nm flash. The figures suggest that Toshiba's BiCS 3D flash is efficient enough to go to QLC (4-bit) levels and still maintain a higher margin than their current MLC (2-bit) 2D flash.

More to follow as we continue our Flash Memory Summit coverage!

 

[cbn]

sandisk did it back in 2009 already
https://www.sandisk.com/about/media...it-x4-(4-bits-per-cell)-nand-flash-technology

So 43nm lithography.

Did they launch QLC on smaller lithography as well? I couldn't find info they did searching for terms "Sandisk X4" or "Sandisk QLC".

 

[Magic Carpet]

Well, laptops are a bit different. You can still buy a quality laptop if you're willing to pay. Dell XPS13 has aluminum shell, carbon fiber palm wrest, minimalistic screen and QHD screen. You can still have quality.

Choice is still there, agreed.

However, I'm not sure if SSDs will follow the same direction. In case of laptops it's easy to see what you're paying big money for. In case of SSDs and reliability, not so much. I'll keep buying MLC for as long as it's available. I value reliability too much to buy TLC.

Same here. I suspect, in a few years we might have to buy Enterprise class products if we still want quality/data/power protection. I also hate TLC drives, not just because of questioned reliability but speed issues as well, they can be very slow at times. Recently tested several models I used for my budget builds (Crucial & Toshiba).

There is still a market for what we want, though. We shall see. Personally, I am all right with high price and low storage, as I store my data mainly on HDDs anyway. Still haven't found a reasonably priced Intel 320 SSD Series alternative replacement for my rigs, though. Not many consumer SSDs have power protection these days.

You don't need see past 768p TN LCDs to see where the rest of the PC industry is going.

Yeah... might get a Microsoft surface device next time. At least these are made of magnesium. I like quality materials. When I was at their Store couple years ago, they let me drop the device on the floor, just to prove the point of how reliable that chassis is..

 

[StrangerGuy]

Yeah, like with the laptops. They used to be quality, now we see lots of $200-300 models made of junk. Same idea.

Hell, even iPhone SE is getting there, lmao (still quality though). Step by step.

You don't need see past 768p TN LCDs to see where the rest of the PC industry is going.

 

[cbn]

Assuming this problem with planar TLC is able to be fixed in the same way it was with the 840 EVO (which involves re-writing the old data to make it fresh again) how much do we know about planar TLC that is left unpowered? (eg, laptop that hasn't been used for a few months or even a year).

Not much, and for me that's the problem. Anything "CPU" or "GPU" gets dissected in minute detail by over a dozen of the usual tech sites. Yet the inability / unwillingness of literally any of the same tech sites to do even basic data retention or longer term performance tests (not just SSD's but USB sticks too) leaves a "deafening silence"...

I also don't see monthly data rewriting as a "fix" that actually solves the problem. At best it's a cheap workaround that hides it by massively reducing (already lower) real-world available P/E cycles (first by having to use an SLC cache to hide TLC's lower native performance whilst larger 500GB-1TB MLC drives write directly in one go), then having to constantly monitor the age of and then rewrite older TLC data due to having far lower overhead between 8 voltage states vs 4). This further widens the MLC vs TLC endurance gap far disproportionately to the barely 20% price difference. Even Anandtech's review technique of constantly writing over a period of days until life left in % drops by 1% then calculating "predicted endurance" hides the issue. And yet again, they don't even touch data retention, powered or unpowered.

Best case : Unpowered data retention isn't an issue. But I'll be more happy when I actually see someone take half a dozen TLC drives, image them with a load of CRC checksummed files, put them in a drawer for 6 months then test both speed & data accuracy. Then wipe clean and repeat with 12 months. SSD's are one of the reasons I've lost all faith in the common sense of modern tech review sites and their "storage experts" seeming unwillingness to do any serious testing beyond copy & pasting PR snippets from the manufacturer.

While those tests would be really nice, how do you expect review sites to actually do tests like that.

I'm not asking for 5 year tests or even the largest capacity of each one, just a more useful 6-12 months test on small 256GB drives given that's the time-scale we saw on the 840's problems. After all, the 850 EVO came out in Dec 2014 and a 6 month test would have been completed 10 months ago so it's hardly as if you'll only know the results after the drive is obsolete. It doesn't have to be done for each SSD either, just give people a general idea of real-world 40nm vs 20nm vs 16nm process retention expectation beyond marketing fluff and an endless string of "probably's".

I would definitely like to do some kind of unpowered data retention test with both a drive that is relatively unused and then one that I have written lots of data to in order to wear it out for reasons I mentioned below:

Also another thing I am wondering about is how much these SM2256 drives slow down when the NAND's tunnel oxide wears out. According to the Anandtech SM2256 article LDPC soft decode (and RAID parity mechanism) kicks in when this happens......but it is slower than the fast decode that is used when the cells are new.

http://www.anandtech.com/show/9165/...-reference-design-ssd-review-tlc-for-everyone

Architecturally the SM2256 shares the same core design as its predecessor SM2246EN. The design is modular, which allows Silicon Motion to change parts of the controller without redoing the rest. It features the same single 32-bit Argonaut RISC processor core as the SM2246EN, which is quite unique because we have seen many SSD controller vendors moving towards multi-core ARM architectures. A single custom core obviously brings efficiency gains and we've witnessed those in the SM2246EN, but the downside of such limited CPU power is sustained performance when the controller has to perform garbage collection at the same time as processing host IOs.

The only dramatic change is in the error correction circuitry as the SM2256 supports Low Density Parity Check (LDPC) error correction codes instead of more common and less powerful BCH ECC. Silicon Motion calls its ECC technology as NANDXtend, and it's a combination of LDPC hard and soft decode along with RAID5-like data recovery. The benefit of having three levels of ECC is performance because LDPC soft decode and recovery from parity both have a relatively noticeable impact on performance and are typically only needed when the drive approaches its end of life (i.e. when the NAND has been cycled a lot). Uncycled NAND has much higher reliability because the tunnel oxide hasn't worn out due to P/E cycles, so only very little ECC is needed and LDPC hard decode is sufficient and also doesn't have a dramatic impact on performance.

The reason why hard decode is faster than soft decode lies in how the voltage of a cell is sensed. Hard sensing is binary based, so for an SLC cell like in the graph above the cell can be either 1 or 0. However, as you can see, the voltage threshold distributions overlap slightly and that's actually far worse with MLC and TLC since there are more voltage states. In soft sensing the voltage distributions are divided into several segments, which requires more precision and iterations. For example in segment 4 the bit value can be either 1 or 0 as the distributions overlap, so probability algorithms are used to figure out the correct value. To be honest, ECC codes and the way they work are way over my head, but in case you are familiar with ECC and want to learn more, I suggest you simply google LDPC as there are numerous publicly available academic papers that go into more depth about this topic.

Silicon Motion claims that its NANDXtend technology can extend the endurance of TLC NAND by up to three times, making TLC more robust for heavier workloads and also allowing the use of lower quality NAND that some OEMs may use anyway due to the lack of in-house binning equipment. Unfortunately I didn't have any time to do extended endurance testing with the SM2256 yet to validate Silicon Motion's claims, but I will be sure to test that once we have a retail drive on our hands

So I am thinking maybe two ADATA SP550 120GB SSDs?

One gets tested in new condition for speed and then gets stored unpowered for at least six months.

And then the second ADATA 120GB SP550 SSD gets written to varying levels of levels of wear to in order to wear down the channel oxide. At each wear interval I would measure speed to determine how much the LPDC soft decode and RAID-5 recovery slows down the drive compared to a new one. Then this one also gets stored unpowered for at least six months to determine how well a worn drive retains data.

Any other suggestions? (I'll probably make an official thread too before I start this experiment. I actually have a lot to read about regarding testing SSDs in general as well)

 

[ShintaiDK]

Yeah, like with the laptops. They used to be quality, now we see lots of $200-300 models made of junk. Same idea.

Hell, even iPhone SE is getting there, lmao (still quality though). Step by step.

Disagree on the SE tho

That one is simply about getting in touch with reality. And you get all the best goodies.

 

[Magic Carpet]

Disagree on the SE tho

That one is simply about getting in touch with reality. And you get all the best goodies.

Have you gone one yet?

Glad, they finally figured there is a big market for 4" devices. So unusual, that they priced it so low, imo. This will certainly eat into the sales of Samsung and other vendors. I am not complaining, though. I'll get one soon myself

I hope, the Aluminium is exactly of the same quality as on the 6S. They strengthened it considerably versus the regular 6.

 

[hojnikb]

So 43nm lithography.

Did they launch QLC on smaller lithography as well? I couldn't find info they did searching for terms "Sandisk X4" or "Sandisk QLC".

nope, because that would just be crazy back then.

 

[VirtualLarry]

http://promotions.newegg.com/intel/16-2590/index.html?icid=357056

"Rev Up Reliability - Intel 540 Series".

Kind of ironic for them to use that tagline for their new TLC SSD series. Yeah, TLC. Rev Up Reliability. Doesn't quite make sense to me.

 

[redzo]

TLC is just slower than MLC. Reliability is directly linked to the product itself. A TLC drive can be more reliable than a MLC one.

TLC has its place in the industry. TLC was designed to bring price/GB down while sacrificing endurance and some speed. If your application requires MLC yet you drop TLC, it means that you're a moron and it has nothing to do with TLC tech being bad.

 

[VirtualLarry]

TLC is just slower than MLC. Reliability is directly linked to the product itself. A TLC drive can be more reliable than a MLC one.

TLC has its place in the industry. TLC was designed to bring price/GB down while sacrificing endurance and some speed. If your application requires MLC yet you drop TLC, it means that you're a moron and it has nothing to do with TLC tech being bad.

What are some (consumer) applications that would "require" MLC? I'm curious. Especially since today's TLC drives seem to be marketed as general-purpose drives. Should they not be?

Edit: I don't see P/E cycles or other reliability measures advertised by SSD vendors, other than what gets reported by the trade press, that most consumers never see.

Like, if I'm at BestBuy, will it be obvious to me, the customer, that the PNY CS1311 TLC SSD is likely to wear out 6X faster than their older CS1211 MLC drive? And am I therefore a "moron" for being oblivious to that fact, since it's not clear disclosed on the packaging? (I'm not sure that the packaging even discloses the MLC versus TLC factor, never mind educates on what that actually means for endurance.)

What if I'm purchasing it for a QuickBooks workgroup server at the office to speed it up? Am I a moron then?

 

[PingSpike]

I'm probably just ignorant on this but I never understood Intel SSDs. Their first ones cost a fortune and were slower than competitors, but I guess they had nice warranties and a SSD was a SSD. Then they stuck the sandforce controllers made infamous by OCZ on them with the same high price making them meh in my eyes for a different reason. Now they follow late again and I can only assume they'll still be expensive. Only reason to buy them is if you want the warranty because they're just the same stuff you were already buying with a higher markup. And you're probably just better off buying the cheap one and coming up with a better backup strategy and self insuring since you should be doing that anyway and if the drive dies you can get a replacement for cheaper tomorrow.

I don't really buy the "Intel reliability" thing. They just rebadge other people's stuff with their name, how does that imply they're the only ones with the secret sauce?

 

[redzo]

What are some (consumer) applications that would "require" MLC? I'm curious. Especially since today's TLC drives seem to be marketed as general-purpose drives. Should they not be?

Edit: I don't see P/E cycles or other reliability measures advertised by SSD vendors, other than what gets reported by the trade press, that most consumers never see.

Like, if I'm at BestBuy, will it be obvious to me, the customer, that the PNY CS1311 TLC SSD is likely to wear out 6X faster than their older CS1211 MLC drive? And am I therefore a "moron" for being oblivious to that fact, since it's not clear disclosed on the packaging? (I'm not sure that the packaging even discloses the MLC versus TLC factor, never mind educates on what that actually means for endurance.)

What if I'm purchasing it for a QuickBooks workgroup server at the office to speed it up? Am I a moron then?

Larry, why did you get the feeling that I called you a moron? You can definitely make your way through some non disclosed TLC based drive marketing stuff right?
http://promotions.newegg.com/intel/1...ml?icid=357056

"Rev Up Reliability - Intel 540 Series".

Kind of ironic for them to use that tagline for their new TLC SSD series. Yeah, TLC. Rev Up Reliability. Doesn't quite make sense to me.

This is wrong judgement as intel is just sayin' that their TLC is better than their competitors as a lot of TLC drives are reliable.
Somehow I get your frustration: that TLC is inferior to MLC and that manufacturers hide the main drawbacks of TLC in their marketing papers. Still, TLC is what you pay for, and any crap TLC drive can fail anytime just as any crap MLC drive. Intel's marketing material is all right.

 

[Topweasel]

Larry, why did you get the feeling that I called you a moron? You can definitely make your way through some non disclosed TLC based drive marketing stuff right?
http://promotions.newegg.com/intel/1...ml?icid=357056

This is wrong judgement as intel is just sayin' that their TLC is better than their competitors as a lot of TLC drives are reliable.
Somehow I get your frustration: that TLC is inferior to MLC and that manufacturers hide the main drawbacks of TLC in their marketing papers. Still, TLC is what you pay for, and any crap TLC drive can fail anytime just as any crap MLC drive. Intel's marketing material is all right.

He is saying as though he is an average consumer at BB. We live and breathe this stuff. But is James Smith walking into best buy going to know a TLC drive from an MLC drive. That while the performance might be close Is he going to know that the TLC drive won't last as long as the MLC drive? Is he going to know that performance can take a nose dive if he doesn't access data enough? Is he going to know that in a powered off state that the MLC drive is going to last much longer before losing data?

 

[VirtualLarry]

I didn't think that you were directly calling me a moron, redzo. Just making the argument from the POV of a BestBuy typical customer.

They might just buy a TLC drive, because of price, newness (newer is better, in technology, right?), whatever (might be the only type of drive available by next year). They won't know the intricate difference between memory types.

I guess, I just feel that the lifespan info should be disclosed on the package. Kind of like how many NITS brightness your LCD screen has.

 

[cbn]

TLC was designed to bring price/GB down while sacrificing endurance and some speed.

Going by the prices in the following thread:

http://forums.anandtech.com/showthread.php?t=2472749

It hasn't done much for reduction of price so far.

 

[Elixer]

TLC is just slower than MLC. Reliability is directly linked to the product itself. A TLC drive can be more reliable than a MLC one.

TLC has its place in the industry. TLC was designed to bring price/GB down while sacrificing endurance and some speed. If your application requires MLC yet you drop TLC, it means that you're a moron and it has nothing to do with TLC tech being bad.

No, that is not the case at all, (planar) TLC is not just slower, and have worse endurance, it also is less reliable.
TLC is inherently more unstable by design than MLC and thus TLC firmware is much more complex than MLC as well.
This is Sammy's pic

I think the graphs speak for themselves.

Yes, (planar) TLC has its place in the industry, at that place is squarely at the lower tier market where price is everything.

3D TLC, jury is still out, though, it does look promising.

 

[cbn]

Yes, (planar) TLC has its place in the industry, at that place is squarely at the lower tier market where price is everything.

I've been thinking if a sale priced 480GB MLC drive with SM2246EN controller like this one is $99,99 free shipping then what could we expect of a sale priced 480GB TLC drive with SM2256 controller? $75?