Toshiba, SanDisk Piloting 3D NAND That Doubles Previous Capacity 61
Lucas123 writes: Under a joint development agreement, Toshiba and SanDisk have begun pilot production of a new 48-layer 256Gb NAND flash chip in a brand new fab in Mie prefecture, Japan. The new X3 chips, which double capacity from 16GB to 32GB over the previous product, are made with triple-level cell (TLC) flash compared with Toshiba's last multi-level cell (MLC) chip, which stored two-bits per transistor. The chips are expected to begin shipping in products next year. The companies plan to use the new memory in a wide number of products, including consumer SSDs, smartphones, tablets, memory cards, and enterprise SSDs for data centers, the companies said.
Re: (Score:2)
It will be a long time before that gets cheap enough to make flash obsolete, if ever. XPoint is going to be priced somewhere between enterprise flash and DRAM, which will put it at around ten times the price of consumer flash.
If you gave me the choice between a 1TB SSD using NAND, or a 100GB SSD that had much more performance and endurance, I would take the 1TB SSD.
Re: (Score:3, Insightful)
XPoint won't be 10x the price - more like 3x~4x the price of consumer flash when it hits the market.
The economy of scales will drive production up as it is incorporated into mobile devices and enterprise systems, which will also drive down price. XPoint is really more a matter of dooming platter drives to extinction, because the durability, power consumption, and speed will make it highly desirable in the enterprise market.
I wouldn't be surprised if XPoint hits the "tipping point" of $200 for 512GB within 2
Re: (Score:1)
A 64GB NAND MLC flash chip trades for $1.60-$3.57. That means 1TB of MLC NAND costs about $57 at most ($25 at the least). This TLC uses half the die space to store the same amount of information--notably, it uses one multi-level transistor to store more data than one of some other type of multi-level transistor--so the production cost will be the same, but the first batches may have production problems cutting off a portion from usefulness, scaling costs (e.g. 50% fail, double the price). That means we c
Re: (Score:2)
64Gb, not 64GB. Multiply your prices by 8.
Still, even at ~$150/tb the use cases for spinning disks are steadily diminishing.
Re: (Score:2)
I think you got bits and bytes confused, pretty sure there's no 64GB NAND = 512 Gbit NAND to be had anywhere. So you can multiply all those prices by 8 for $205-457 for a TB. Even if it goes down to $160/TB that's still a lot more than HDDs.
Re:Too bad (Score:4, Interesting)
What HDD makers really need to do is stop focusing on price and make a line of drives that is made to be archival grade. For example, there was a line of drives with two read/write heads that worked in an active/active configuration.
What might be even better would be to make a standardized, rugged drive cartridge case, similar to iMation's RDX. Something that can handle drops, be gripped easily by a tape silo's robotic mechanism, can handle tens of thousands of mounts/dismounts, has built in encryption, the ability to have WORM functionality (similar to late gen DLT drives where the cartridge can be formatted as normal or WORM), and so on. The drive can be presented either as a tape volume, standalone JBOD hard disk, or part of a RAID set (and inserted/ejected at the same time with 2-3 companions.)
Moving HDD to a backup/archive use as opposed to primary storage will keep this technology relevant, as opposed to trying to fight with SSD (which is a better primary storage technology [1].)
[1]: In all ways but recovery. An SSD goes bad, there is no way to recover the data, period.
Re: (Score:2)
Normal people aren't going to mess with a dead drive and serious enterprise customers are not going to have only one copy of their data.
Ultimately, all that matters is cost. Speed is even somewhat of an optional thing with the larger archival volumes. Although you do need enough speed to make populating (or recovering from) an entire drive practical.
I'm recovering a 4TB drive right now and it's moving as fast as all of the associated bottlenecks will allow.
I don't care what the underlying tech is. It just c
Baffled by troll moderation (Score:2)
I genuinely like to know what is "trollish" about my post. I'm just trying to make the reasonable prediction that cheap/dense SSDs and XPoint mean more about the death of platter drives as a storage medium than XPoint making SSDs obsolete.
I also agree with the other point made here that HDD manufacturers would be better served at looking to be a future replacement for tape media as an enterprise archival method.
Perhaps my mention of the artificial propping up of prices angered some slashdot mods? It's not l
Re: (Score:2)
The general idea that SSDs will gradually encroach on HDD markets is correct and has been happening, albeit at a somewhat slow pace. While predictions about the disappearance of HDDs have existed for almost the last ten years, units sales of HDDs still dominate and probably will for quite a few more years. Microdrives disappeared about a decade ago, but that has so far been the only HDD market to completely evaporate. SSD are increasingly gaining ground in low-margin markets, such as laptops, but even th
Re: (Score:2)
So far SSDs have only barely kept up with software bloat. They need more density.
Re: (Score:2)
The HDD manufacturers are still making the 15k RPM SAS HDD. Why? Because you can't use cheap SSDs in the data center. So the enterprise SSDs we're talking about are several times more expensive than the laptop SSDs. Depending on the workload, cost per performance may or may not be better for SSDs. Because SSDs are mostly considered as a caching layer between DRAM and HDDs, the workloads tend be be write-heavy, which increases wear leveling overhead and impacts performance.
So, will SSDs kill off the ent
Re: (Score:2)
XPoint 3D still has a ways to go with price (and the fact that it isn't out in the field yet.) It is still too expensive to be a 100% replacement for SSD, just like SSD is too expensive to replace HDDs everywhere.
However, XPoint 3D does have its uses. Loading the core OS, application, and kernel come to mind as well as having a swap volume (pretty much the same concept as mainframe "external RAM" which was slower.)
Re: (Score:2)
Or in a hybrid platter/XP3 or SSD/XP3 drive, like platter/SSD hybrid drives now.
Re: (Score:2)
Maybe, maybe not. Detailed specification of a xpoint 3d final component has still to exists, while the 3D NAND is the continuity of a well know technology. I am particularly curious at the temperature effect on the xpoint 3d cells, because others "innovative" memory that exists today are not very good in this area.
Re: (Score:2)
XPoint is less dense. A lot less dense. So it will be more expensive for GB.
Will not buy TLC NAND (Score:3, Insightful)
Re: (Score:3)
You know I thought you were wrong and that a Triple Level Cell had 3 levels per cell, storing about 1.6 bits. Apparently I was wrong. TLC despite standing for triple level cell apparently means in fact 8 levels per cell where as MLC (multi-level) is for 4 levels per cell.
Whoever came up with these terms is nuts.
Re: (Score:3)
Re: (Score:2)
Makes perfect sense to me. Maybe its because I understand binary.
How so? The cells have multiple voltage levels, they don't store binary. 3 levels is not the same as three bits.
Re: (Score:2)
3 levels is not the same as three bits.
You had posted that you just discovered that it is, and now you are posting that it isn't. Make up your mind.
AC poster is right. You must have thought that SLC stored 0 bits since it had only 1 possible "level."
Re: (Score:3)
You had posted that you just discovered that it is, and now you are posting that it isn't.
Try actually reading what I wrote before being overcome with smugness. MLC can measure multiple (i.e. > 2) voltage levels on the floating gate transistor. The number of bits is the log_2 of the number of distinct voltage levels.
Calling something with 3 BITS that has 8 LEVELS triple level is silly. Why not just call it 3 bit cell. SLC sort of made sense from an EE point of view as being able to store one level as di
Re: (Score:2)
The level of ignorance on this thread is truly staggering:
If 1.6 bits per cell ever made any sense to you, then you should just leave the discussion.
Well, indeed. It seems I'm trying to talk technical things with people who don't even have the most elementrary grounding in information theory.
Re: (Score:2)
Calling something with 3 BITS that has 8 LEVELS triple level is silly.
Especially since it only has 7 threshold levels... which was clearly explained to you hours before you made this reply. Typo, or dumbo?
Re: (Score:2)
Especially since it only has 7 threshold levels... which was clearly explained to you hours before you made this reply.
One threshold level = 2 distinguishable voltage levels. 7 thresholds = 8 distinguishable voltage levels. 3 bits is still not 3 levels by any normal definition of the word.
Typo, or dumbo?
It's entertaining that you're saying such things after making such elementary mistakes.
Re:Will not buy TLC NAND (Score:4, Funny)
Did you think SLC (single level cells) had 1 levels per cell, thereby storing 0 bits?
I own one of those. It reads zero in all locations.
Re: (Score:2)
And Wikipedia goes into more detail: https://en.wikipedia.org/wiki/... [wikipedia.org]
SLC - Single Level Cell = 1 bit (2 states), most robust
MLC - Multi Level Cell = (typically) 2 bits (4 states), ~1/10 of the lifespan of SLC
TLC - Triple Level Cell = 3 bits (8 states), ~1/10 of the lifespan of MLC
Re: (Score:3)
Well, smaller process sizes also reduce write cycles. When they went from 2D to 3D they went back to a bigger process size with less defects due to all the layers. Though they had some controller/firmware issues the first 3D TLC NAND had more raw write cycles than state of the art planar MLC NAND. Of course now they're shrinking it again in the quest for even more storage, but the clock got a pretty good reset going from 1 layer to 48. Going from MLC to TLC is more of a variation that cuts write cycles to a
Re: (Score:2)
For pete's sake, MLC stands for multi-level cell, not four-state (two bit) cell. TLC is just a kind of MLC; it's not an either-or. TLC is eight-state (three bit). I realize we're kind of stuck with dumbed-down nomenclature, where MLC is used specifically for four-stat
Re: (Score:2)
Back in my day (Score:3, Funny)
Our storage mediums spun and made noises! And we liked it!
Re: (Score:2)
*cups ear* Whut?!
Re: (Score:2)
Back in my dad's day the memory spun and made noises [wikipedia.org]. No idea if he liked it.
Re: (Score:2)
Back in my dad's day the memory spun and made noises [wikipedia.org]. No idea if he liked it.
Back in my day I was an operator on a Burroughs B3500. This machine had core memory. If you bumped the tape cart against the memory cabinet a little too hard it would cause a memory fault. Apparently those ferrite donuts didn't like being jostled.
Re: (Score:3)
Our storage mediums spun and made noises! And we liked it!
What is funny is that I recently upgraded my laptop to an SSD and I was flabbergasted that during disk access it made the exact same noise as it did when I had a regular hard disk. I thought I had been sold a fake SSD, but I’m getting approx 510MB/s on benchmarks, so I know it is legit. On further investigation the noise comes from where the speakers are located, so what I thought was hard disk noise was probably just electrical noise. My system makes the same soft buzzing sound on data access as it a
X3 (Score:1)
Micron is ahead (Score:1)
Re: (Score:2)
Actually Samsung is ahead because they are the only ones selling 3D NAND right now. The Micron press release is more impressive than Toshiba's, sure, but it will take a couple of months before it gets into store shelves.
Re: (Score:1)
Re: (Score:2)
Why is it that the video at the link you sent compares the revolution to the jump from a single-story office to a 32-story highrise, yet, as a result, memory is only 3x as much? I'd expect a multiple of 32, or even if the old stuff was TLC and the new is SLC, a multiple of almost 10.
Re: (Score:1)
Re: (Score:2)
Holo storage was supposed to be out back in 1991-1992 (Tamarak), then about 10 years later, InPhase supposedly had a drive for it, but never made it to the market (IIRC).
Would be nice if that technology would get off the ground, but so far, it has been nothing but vapor. I would wager Half Life 3 comes out well before then.
Re: (Score:2)
http://akoniaholographics.com/products/ [akoniaholographics.com]
they switched from offering a end user drive to full rack-mounted usage scenario after they went bankrupt, and changed their name.