TR Forums

HERETIC · Wed Dec 14, 2016 1:43 am

Are we going to see a reduction of SSD manufacturers and increases in price?
Reason I ask is this article alarmed me a little-
http://www.anandtech.com/show/10903/mic ... production

This paragraph-
"Moreover, Micron seems to supply 3D NAND to only one third party vendor now, ADATA, an indicator that the company is gradually leaving the spot/contract market of NAND flash chips. Note that this does not mean that Micron will never sell NAND to any third party at all, because a commercial company is supposed to make money. Keeping in mind that Toshiba/SanDisk as well as SK Hynix also do not supply SSD-grade 3D NAND to third parties, ADATA remains the only independent maker of actual drives that ships products featuring 3D NAND memory."

seems to point to a possible monopoly by the six manufacturers of Nand and a select few.

Arclight · Wed Dec 14, 2016 3:19 am

A monopoly of 6 is not a monopoly unless one of them manufactures way more than all of the rest or there is a conspiracy, an unofficial deal among them not to cut prices and just settle which one gets what segment of the market and never really compete. Personally I can't imagine even that stopping price drops long term just because of manufacturing. It's normal that there would be only a few companies left standing eventually, that happened in the CPU and GPU world, why wouldn't it happen for storage? Let's not forget that before SSDs took over the market by storm, there weren't that many HDD manufacturers either and yet prices were acceptable.

If anything, the bigger issue is if NAND can continue to be the performance type of permanent storage. There are so many alternatives being worked on to retire this technology that it's unlikely to survive the next decade. Then there is the migration from the current materials and manufacturing processes to next generation ones in order to escape the death of Moore's Law. While it might be possible to get close to 1nm manufacturing, it's impossible to continue afterwards.

Wed Dec 14, 2016 5:44 am

Arclight wrote:
A monopoly of 6 is not a monopoly

https://en.wikipedia.org/wiki/Oligopoly

TravelMug · Wed Dec 14, 2016 7:45 am

Prices on SSD-s are already going up. The reason for now is plausible to be shortage, but realistically it does not help that there is only a handful of NAND manufacturers left. As JAE pointed out, it's not a monopoly, but an oligopoly and that is the same bad, just with a bit more effort regarding coordination on their part. It's actually worse for price fixing because it is harder to detect and prove compared to a monopoly.

As for SSD prices going down anytime soon, I would not count on it. Same with RAM and HDDs. This has been in the making so to say for month now, I like this summary article about it from a week ot two ago because it is a nice recent refresh with current data:

SSD Prices Skyrocket As NAND Shortage Deepens, HDD Shortage Looms As Components Become Scarce

Chrispy_ · Wed Dec 14, 2016 8:27 am

WHOA WHOA WHOA WHOA NOW!!
You're saying there might be a cartel of memory producers including Hynix, Micron, and Samsung among others?
That would never happen.

Glorious · Wed Dec 14, 2016 8:56 am

TravelMug wrote:
The reason for now is plausible to be shortage, but realistically it does not help that there is only a handful of NAND manufacturers left. As JAE pointed out, it's not a monopoly, but an oligopoly and that is the same bad, just with a bit more effort regarding coordination on their part. It's actually worse for price fixing because it is harder to detect and prove compared to a monopoly.

"Shortage", used loosely as it is in this sense, is exactly how most (if not basically all) price-fixing is actually accomplished: You indirectly fix the price by directly setting the production. That is, they collude towards production quotas, not so much stated prices.

There are numerous reasons for this. One is that "fixing the price" while not modifying production doesn't really make money (what do you do with the excess production if you are pushing the price up?). Another is that most of their buyers have long-term relationships with contracts, not so much spot market. Thus, on re-negotiation, constrained supply forces both sides to settle on a higher price: Your sales guy can't cheat the cheating conspiracy by deviating slightly from the pricing agreement, as he doesn't actually have the units to move. In fact, he doesn't have to even know about it. (Always a good thing in illegal conspiracies!)

In short, the stated reason for the price increase is essentially indistinguishable from the mechanism of price fixing. Hence, whether or not it is "plausible" doesn't really have any relevance to whether or not price fixing is occurring.

As you said, this is "actually worse" than monopoly because monopoly is far easier to detect!

EDIT: Is price fixing occurring? As Chrispy_ said and cited, lol, it's not remotely unlikely and *I* certainly wouldn't bet one cent against the possibility.

TravelMug · Wed Dec 14, 2016 9:31 am

Oh yeah, there is definitely something going on in the background as well, but in this case I think it is also mixed with a healthy dose of messing up forecasts and some other selfinflicted problems. The fact that this is not predicted to stabilize for at least a year is already the price fixing territory.

It is also "funny" how promiscuous the situation is, it's like a box of hamsters. The same names pop up in all categories where the price hikes are happening.

Samsung for RAM, SSD and HDD
Toshiba for HDD and SSD
WD for HDD and SSD
Micron for SSD and RAM
Seagate for HDD and SSD

And for some geeky standup routine you could use the fact that this whole SSD "shortage" is happening right after Seagate and WD entered the SSD market through Sandisk and Toshiba

Almost like they got hooked on the sweet returns of the "Taiwan flood HDD shortage and price hike" and needed to continue the action. And I guess Samsung and Micron weren't against when their sweet RAM pricing joyride ended a year or two ago.

meerkt · Wed Dec 14, 2016 9:03 pm

TravelMug wrote:
Seagate and WD entered the SSD market through Sandisk and Toshiba

WD and WD you mean.

HERETIC · Thu Dec 15, 2016 12:56 am

Thanks guys--oligopoly-the word my brain couldn't find.

Back on topic-If you read the Micron blurb-The idea to not sell flash dies, only finished products-
would be good for profits if it worked.
But I also wonder if this is also a punitive knee-jerk reaction to no-one buying their planer TLC.
It was so bad, even their manufacturing partner Intel wouldn't go near it.

We'll only know once Toshiba and Hynix get their 3D out there in mainstream products.

If they took a halfway point and decided to only sell cut and binned dies and not wafers
to customers I'd be behind that 100%-Remembering that's what helped kill OCZ (low quality
nand ending up in SSD's which resulted in a 60% failure rate.)

strangerguy · Sat Dec 17, 2016 9:53 pm

My speculation is Apple raking up obscene per-unit profits with the iPhone has made everybody in the electronics business heavily questioning their usual strategy of moving product at razor-thin margins for the sake of volume. I bet everybody would rather sell at a price hike of 1.5x per unit and losing like 20% volume in the process because it actually earns more money.

Kougar · Wed Dec 21, 2016 11:23 am

"ADATA remains the only independent maker of actual drives that ships products featuring 3D NAND memory."

That quote can be misleading. Transcend still ships its 370S drives using both Micron and SanDisk 15/16nm MLC NAND, but it's 2D and its endurance ratings reflect that.

The number of NAND suppliers has always been shrinking and the number of players in the SSD space along with it. But it would be a big gamble for Micron to only sell NAND to Intel and hold the rest for its own drives. They have to produce a minimum amount of NAND at each facility to offset fixed costs of that brand new fab and their older fabs so Micron can only go so far with this sort of strategy.

The end of the article noting 256GB drives will be the next performance-plagued models on Gen 2 3D NAND is particularly interesting, not everyone needs a 512GB drive and the cost of paying for that unneeded capacity is already going to offset the savings from the Gen 2 stuff. I also want to know what the endurance ratings of the Gen 2 NAND will be.... a Transcend 2D 16nm Micron NAND 1TB drive has a 1160TB endurance rating, and an equiv size 960 Pro is at 800TB. The EVO models already halve that to 400TB.

Vhalidictes · Wed Dec 21, 2016 11:28 am

strangerguy wrote:
My speculation is Apple raking up obscene per-unit profits with the iPhone has made everybody in the electronics business heavily questioning their usual strategy of moving product at razor-thin margins for the sake of volume. I bet everybody would rather sell at a price hike of 1.5x per unit and losing like 20% volume in the process because it actually earns more money.

This would have been a good strategy over the past decade, but now that people are seriously cutting back on hardware purchases and/or reducing disposable income that might not be a good idea.

I guess we'll see, but I'm sure that people will be able to put up with HDD's for a lot longer than baseline if the SSD market becomes too pricey.

The Egg · Wed Dec 21, 2016 1:19 pm

Speaking of which, why are the least expensive 256GB SD cards more expensive than an equal size SSD? Chip density? With no meaningful controller and likely inferior NAND quality, you'd expect them to be less.

Waco · Wed Dec 21, 2016 1:30 pm

The Egg wrote:
Speaking of which, why are the least expensive 256GB SD cards more expensive than an equal size SSD? Chip density? With no meaningful controller and likely inferior NAND quality, you'd expect them to be less.

They're the densest dies available, so they're a premium product. At least, that's my understanding of it.

Wed Dec 21, 2016 9:46 pm

Waco wrote:
The Egg wrote:
Speaking of which, why are the least expensive 256GB SD cards more expensive than an equal size SSD? Chip density? With no meaningful controller and likely inferior NAND quality, you'd expect them to be less.

They're the densest dies available, so they're a premium product. At least, that's my understanding of it.

Densest and probably cherry-picked as well (for fast write speed and low power consumption).

Waco · Wed Dec 21, 2016 11:36 pm

Also, just my opinion, but prices aren't going anywhere but down as more fabs come online and fabs improve their processes.

DancinJack · Wed Dec 21, 2016 11:49 pm

Waco wrote:
Also, just my opinion, but prices aren't going anywhere but down as more fabs come online and fabs improve their processes.

While I do mostly agree with that sentiment, there is a NAND shortage going on right now that is driving prices up. So short term, up, and long term, down.

HERETIC · Thu Dec 22, 2016 1:56 am

strangerguy wrote:
My speculation is Apple raking up obscene per-unit profits with the iPhone has made everybody in the electronics business heavily questioning their usual strategy of moving product at razor-thin margins for the sake of volume. I bet everybody would rather sell at a price hike of 1.5x per unit and losing like 20% volume in the process because it actually earns more money.

Problem is "everybody" is not capable of pulling off the "big con" that apple has spent years mastering.................................

HERETIC · Thu Dec 22, 2016 2:09 am

The Egg wrote:
Speaking of which, why are the least expensive 256GB SD cards more expensive than an equal size SSD? Chip density? With no meaningful controller and likely inferior NAND quality, you'd expect them to be less.

Simple answer-Because they can.
SD cards follow other storage-Example HDD's-3 to 4TB is the sweet spot at the moment-The further up you go the more it costs per TB.
apple used to be the master at this-add $6 worth of flash-charge $100 extra...................................

Kougar · Thu Dec 22, 2016 6:07 am

It's density and design complexity. Same exact reasons the same desktop hardware still costs more when inside a laptop, taken to an entirely new level.

Architecting a single thumbnail sized NAND chip to have up to 275MB/s sequential performance is no small feat and requires design complexity. It is not an SSD where they have multiple chips and channels to extract parallelism from.

By comparison it's kind of amazing that a UHS class 3 card can attain 275MB/s reads and 100MB/s writes when the first generation X-25 SSD had 10 chips and 10 channels to work with and still couldn't reach that.

The Egg · Thu Dec 22, 2016 8:39 am

HERETIC wrote:
The Egg wrote:
Speaking of which, why are the least expensive 256GB SD cards more expensive than an equal size SSD? Chip density? With no meaningful controller and likely inferior NAND quality, you'd expect them to be less.

Simple answer-Because they can.
SD cards follow other storage-Example HDD's-3 to 4TB is the sweet spot at the moment-The further up you go the more it costs per TB.

Sounds about right. I'm going to be modding my 2009 iPod Classic with one of these. I had initially planned to just get the dual full-SD design (1x256, 1x128), but found it was less expensive to get the 4x MicroSD model and fill it with more chips at smaller densities. 2x128 is less expensive than a single 256, and if you can get down to 64GB cards, you really save.

I ended up ordering 2x128 and 2x64 for a total of 384GB, and saved about 30 bucks.

HERETIC · Thu Dec 22, 2016 6:07 pm

Step in the right direction-
http://www.anandtech.com/show/10927/gal ... th-3d-nand
Tho I tend to see Phison a bit like the old OCZ/Sandforce-"Customers as crash test dummies"

TR Forums

SSD prices and Competition.

SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Re: SSD prices and Competition.

Who is online