Question How much will Samsung 970 Evo 2TB cost a year from now?

[anandtechreader]

Hello, I read the rumors that SSD cost will drop this year. How much will the Samsung 970 Evo 2TB likely be costed a year from now?

 

[cbn]

It could end up being more expensive.

 

[anandtechreader]

How come? There seems to be some predictions that SSD will be cheaper this year. I wonder if the price would be dropped by half.

 

[cbn]

One good thing for 2020 NAND Supply is that Yangtze Memory technologies will be scaling up:

https://www.anandtech.com/show/13752/anandtech-year-in-review-2018-ssds

A New Competitor On The Horizon: Yangtze Memory Technologies Co.

China's Tsinghua Unigroup has a subsidiary called Yangtze Memory Technologies Co. (YMTC) that is trying to break into the 3D NAND market. As a late entrant to the market, their roadmap is significantly behind the competition and requires them to pull off a series of very fast paced technology iterations to catch up. They developed 32-layer 3D NAND in 2017 and are currently sampling 64-layer 3D NAND with mass production planned for late 2019. From there they plan to skip the 96L node and jump to 128L in 2020 to catch up with the more established players. The key difference that sets YMTC's NAND apart is a novel manufacturing method they branded Xtacking: rather than place peripheral control circuitry under the flash memory array (as first implemented by Intel and Micron, and on the roadmap for everyone else), YMTC fabs the two parts of the chip on entirely separate wafers. They claim to be able to bond together finished peripheral and array wafers in a single process step. Their 64-layer 3D NAND is the first demonstration of this technology.


Aside from allowing for a competitively small overall die size, YMTC's Xtacking design has a few other consequences that will help them catch up. By fabricating the peripheral circuitry and memory array separately, YMTC is able to somewhat decouple the development of the two designs and iterate a bit more quickly. The peripheral circuitry can also be fabbed on a traditional logic process instead of a memory process—they're currently using a mature and thus very cheap 180nm logic process. YMTC is planning to make extremely high IO speeds a hallmark of their 3D NAND through the use of many planes per die. They are hoping to reach 3Gb/s from a single die, while the existing 3D NAND players are just starting to move beyond 1Gb/s.

However, I am not sure how much overall effect they will have?

 

[arandomguy]

NAND bit cost are currently forecasted to fall by up to 30% in 2019 by comparison they fell 50% this year. Some of this is going to be from QLC production however. Keep in mind NAND cost is not the only cost and costs are not the only factor in pricing. For example the 970 EVO 2GB costs $170 more than the 860 EVO 2TB, how much of that is cost factors?

The biggest possible disruption for NVMe drive prices would be if Micron/Crucial actually enter a performance variant in the consumer market similar to the MX500 for SATA.

Possible headwinds could arise due to the current trade disputes, QLC switchover pushing up TLC prices, or production throttling... I mean "work place fire."

 

[Billy Tallis]

There will be more price pressure on Samsung's 2TB NVMe drives by the end of the year because there will be many more competitors in that space. I think Samsung's price will end up under $400, and competing 2TB TLC NVMe drives might hit $350 during holiday sales.

 

[Harry_Wild]

The 970 EVO will be cheaper when the faster, better 980 EVO comes out! Just wait till then.

 

[anandtechreader]

During Boxing Day, I bought one 1TB and two 2TB SSDs for my laptop and workstation.

Samsung 970 EVO 1TB at: $277
Samsung 970 EVO 2TB at: $538

I plan to use one 2TB for Windows in my workstation. As for Linux, I can install it in the 2nd 2TB SSD or return the 2TB and use the 1TB for now and then upgrade to 2TB a year later. I wonder if using the 1TB now and upgrade a year later could save some money. My analysis is that even I could save some money by upgrading to 2TB later, I won't save that much unless the 2TB SSD drop a lot a year later.

 

[cbn]

NAND bit cost are currently forecasted to fall by up to 30% in 2019 by comparison they fell 50% this year. Some of this is going to be from QLC production however.

If cost per bit were to drop 30%, but some of this were due to QLC would imply 3D TLC actually increases in price slightly.

 

[arandomguy]

During Boxing Day, I bought one 1TB and two 2TB SSDs for my laptop and workstation.

Samsung 970 EVO 1TB at: $277
Samsung 970 EVO 2TB at: $538

I plan to use one 2TB for Windows in my workstation. As for Linux, I can install it in the 2nd 2TB SSD or return the 2TB and use the 1TB for now and then upgrade to 2TB a year later. I wonder if using the 1TB now and upgrade a year later could save some money. My analysis is that even I could save some money by upgrading to 2TB later, I won't save that much unless the 2TB SSD drop a lot a year later.

The last little while was an anomaly in NAND prices so far historically in which we saw the price spike and stagnation followed by a huge collapse during end of this year. Price fall will likely slow and but still fall consistently as we return to the norm. Prior to the last little bit the general trend was roughly half the price or 2x the capacity over 2 years.

So unless you can actually use the extra space now the price will come down .

The only pause I have with this is QLC. I'm not sure how you feel about it but currently I'd prefer to buy in before there is a push (should it happen) of QLC moving to the mainstream and pushing TLC upstream. The problem isn't those endurance figures that get advertised and fresh drive benchmarks (both misleading in my opinion) but longevity issues due to how much lower tolerance is required. With TLC we had several significant developments to offset this (eventually), 3D NAND, charge trap, and newer ECC (BCH to LDPC). I'm not aware of anything similar on the horizon with the move to QLC.

And the consumer cost savings of QLC might be questionable when it's all said and done. Although with you looking at larger capacities that might factor in more.

Not the mention the possibility of layer stagnation at 128 and once again the need to aggressively target smaller process geometry to increase density.

If cost per bit were to drop 30%, but some of this were due to QLC would imply 3D TLC actually increases in price slightly.

I wouldn't think so because QLC is extremely unlikely to command that large of a market share to off set all other gains by end of this year. Maybe in the very long majority productions shifts to QLC and even TLC gets pushed up to "premium" but that won't be happening in a years span no matter how aggressively they ramp.

It's more that TLC prices won't drop as much as they would if everything were just going into TLC production and QLC was a non factor.

The other thing is Samsung has a lot of mindshare in the SSD space. Some of the expected pressure is expected to also come from Chinese NAND I believe. But how much pressure will that put on Samsung consumer drive prices? Will the end user consider lower reputation manufacturers (eg. TEAM) drives using Chinese NAND?

 

[anandtechreader]

Thanks. I don’t know if I will use up 1TB in a year. Since there are only two NVMe slots in most motherboard, I cannot think of other use of the 1TB ssd if I do upgrade to 2TB later.

As it is for a work machine, reliability is more important than some small saving in price. In this case, is it better to keep the 2TB and return that 1TB rather than the other way around?