Well done Intel and Micron. If these claims are borne out in real, shipping devices it will do to current SSDs what they did to HDs (i.e. leave them in the dust).
It's amazing what they can do when their CPU business is gone to shit...
That's probably what it's about, partially. If they can increase storage speeds enough that CPU speed once again becomes the bottleneck in the system, people will start buying CPUs again.
Nice to see the lack of EE and Physics holders posting on how awesome this breakthrough is when it's a rip off of a memristor and not that impressive.
As an EE, this is massively cool and amazingly impressive unless you think talking about theory of what might be achievable is more impressive than actually putting that theory to practical use and delivering a real product.
This memory technology with superior read/write and number of time write operations is way above what currently being offered. IT will kill Samsung and others in NAND memory space.
On BBC they site it will still be significantly cheaper then DRAM. Let just say it will be half the price..... ( I think significantly means even lower, but i assume 2x cheaper for now. )
DRAM is trading to $2.5 per 4 Gb (512MB), going as low as $2.
3D Xpoint could be $2.5 per 8 Gb ( 1GB ), a 30GB would cost $75, and retailing for $150 or $200. For something as revolutionary as this, it isn't expensive at all.
But before we get ahead of ourselves, this essentially mean the whole stack of storage, RAM needs to be rethink and retested.
Should Xpoint lives as a buffer on SSD? Or better yet lives with DRAM, with DRAM moving closer to CPU. ( Stacked DRAM much like HBM in GPU )
Yes this is exciting, but it would properly be years before we get the full potential.
P.S - And it might turns out to be a perfect fit for Mobile Phones, When you dont want the constant refresh of memory dragging battery,
Also the more granular reads and writes should reduce battery requirements. Something like AppleWatch 2 could use this instead of both RAM and Flash!
This is freakin' awesome. The performance gains from combining RAM/STORAGE is very significant. Data access will be instant. It seems that the latency/refresh rate of the monitor could become the bottleneck which is weird to think about.
This memory technology with superior read/write and number of time write operations is way above what currently being offered. IT will kill Samsung and others in NAND memory space.
Doubtful, since Sammy will just fire up their photocopiers...
If Intel can get this technology to market in a timely fashion, they will have something to drive profits for a short time. As for the technology taking over the msrket, it is highly doubtful. Nantero has a carbon nanotube solution that looks even better than the Intel/Micron product. It's interesting to see Intel coming out with the announcement but hasn't released a product. It seems par for the course for them. Apple actually released a 64 bit ARM CPU as their announcement. No thanks Intel, I'll be purchasing the Nantero NVM product.
Sounds very cool. Once the production costs go down this will spell the end of disk drives.
My main question is how long until Samsung tries to steal the technology and try and sell it themselves.
Unless it is a standard that can be bought from others the most companies will not use it. Micron and Intel will need too license otherwise it will not be cost effective.
Sounds a bit vapor-ware'ish to me. If Intel rolls out SSD's themselves to put the proof in the pudding, then maybe we will see something cost-competitive with NAND flash and finally see an end to mechanical hard drives.
The reason that mechanical drives are still being produced at all is because NAND wears out, quickly. TLC has 10,000 writes, MLC has 100,000 writes, and SLC has 1,000,000 writes. NAND also loses it's charge over time, so without being refreshed every so often, eventually you end up with a "blank brick", not a "read-only" one as some people have thought. So for archival use, NAND isn't good at all.
Xpoint might be able to replace NAND on the M2 PCIe storage, but it seems a bit silly to make SATA3 SSD's with it if it is indeed 1000 times faster.
One quick correction, MCL erase cycles is around 150 to 1500 depending on the lithography and SLC is between 10,000 and 100,000. You data is a number of years old. As you pointed out HDD are here for a while since you can write and erase millions of times and come back to the drives years later and your data is still there no memory solution can do things today.
This is freakin' awesome. The performance gains from combining RAM/STORAGE is very significant. Data access will be instant. It seems that the latency/refresh rate of the monitor could become the bottleneck which is weird to think about.
The list of possibilities is endless.
Hmm ...
If so, the same tech could be use to hold the pixels of the display. My iMac 5K has a resolution of:
Display Type: Retina LCD
Resolution: 5120 x 2880 Retina
Retina: Yes
Pixel Depth: CGSThirtyBitColor
Main Display: Yes
Mirror: Off
Online: Yes
Built-In: Yes
So, if my math is correct, that's 5120 x 2880 x 30
That's ~55 MB of Display storage that can be refreshed anywhere from a single pixel color (out of 30) on up to whatever ...
For those wondering about costs, a 1TB HDD which is the lowest cost since it is a single platter drive is at $0.05/GB and MLC NAND is running around $0.50/GB. So they have to beat this in order to replace it.
For those wondering about costs, a 1TB HDD which is the lowest cost since it is a single platter drive is at $0.05/GB and MLC NAND is running around $0.50/GB. So they have to beat this in order to replace it.
What if they could offset the additional cost with less battery, smaller size, replace SRAM/DRAM/Flash at acceptable performance.
It could really affect current wearables like the Apple Watch -- and open up whole new possibilities for future wearables.
The more I think about it, this tech has some real possibilities in: [LIST] [*] low-memory where speed is a consideration -- AppleTV, Routers, HomeKit Controllers and accessories, SmartTVs, etc. [*] low memory where speed is not a consideration -- appliances, such as refrigerators, washer/dryers, etc. [/LIST]
For the latter, cost would be the major issue -- as the memory would last longer than the appliance.
Comments
This could / will be a real game changer in the industry and in our pockets.
Solid state storage with read-write speeds way over current RAM abilities?
We are talking here about speed of 200GB/s ( SSD ~200MB/s x 1000 ) where DDR3 speed ~17GB/s .. pinch me
Well done Intel and Micron. If these claims are borne out in real, shipping devices it will do to current SSDs what they did to HDs (i.e. leave them in the dust).
It's amazing what they can do when their CPU business is gone to shit...
That's probably what it's about, partially. If they can increase storage speeds enough that CPU speed once again becomes the bottleneck in the system, people will start buying CPUs again.
As an EE, this is massively cool and amazingly impressive unless you think talking about theory of what might be achievable is more impressive than actually putting that theory to practical use and delivering a real product.
Good writeup -- thx for the post.
Maybe this is the 3D SRAM that Micron was rumored to be developing.
I wonder how/if this tech complements or competes with the tech Apple got when they acquired Anobit?
http://www.anandtech.com/show/5258/apple-acquires-anobit-bringing-nand-endurance-technology-inhouse
I assume that Apple is on board with this new tech as both Micron and Intel are major Apple suppliers.
Also the more granular reads and writes should reduce battery requirements. Something like AppleWatch 2 could use this instead of both RAM and Flash!
The list of possibilities is endless.
Doubtful, since Sammy will just fire up their photocopiers...
Unless it is a standard that can be bought from others the most companies will not use it. Micron and Intel will need too license otherwise it will not be cost effective.
They make the chips in my PS4 and the discrete video of Macs. They're even the CPU of your precious Xbox One. so yeah, I've seen them lately.
One quick correction, MCL erase cycles is around 150 to 1500 depending on the lithography and SLC is between 10,000 and 100,000. You data is a number of years old. As you pointed out HDD are here for a while since you can write and erase millions of times and come back to the drives years later and your data is still there no memory solution can do things today.
Hmm ...
If so, the same tech could be use to hold the pixels of the display. My iMac 5K has a resolution of:
Display Type: Retina LCD
Resolution: 5120 x 2880 Retina
Retina: Yes
Pixel Depth: CGSThirtyBitColor
Main Display: Yes
Mirror: Off
Online: Yes
Built-In: Yes
So, if my math is correct, that's 5120 x 2880 x 30
That's ~55 MB of Display storage that can be refreshed anywhere from a single pixel color (out of 30) on up to whatever ...
Nice to see the lack of EE and Physics holders posting on how awesome this breakthrough is when it's a rip off of a memristor and not that impressive.
Certainly respect your pedigree - and I'm not trying to be contentious.
I hold no degrees in either "EE" (electrical engineering???) or physics,
but I'm hoping you'll elaborate your comment.
This may well be an extension (ripoff?) of previous theory, but...
in what way is it not impressive, actually to put it to use after 44 years?
....or, sorry - did I misunderstand? Were you being mildly sarcastic, and saying this shouldn't be
snobbishly dissed on those grounds...?
What if they could offset the additional cost with less battery, smaller size, replace SRAM/DRAM/Flash at acceptable performance.
It could really affect current wearables like the Apple Watch -- and open up whole new possibilities for future wearables.
The more I think about it, this tech has some real possibilities in:
[LIST]
[*] low-memory where speed is a consideration -- AppleTV, Routers, HomeKit Controllers and accessories, SmartTVs, etc.
[*] low memory where speed is not a consideration -- appliances, such as refrigerators, washer/dryers, etc.
[/LIST]
For the latter, cost would be the major issue -- as the memory would last longer than the appliance.
[URL=http://www.patentlyapple.com/patently-apple/2015/07/intel-micron-introduce-3d-xpoint-memory-that-will-advance-8k-gaming-siri-hand-gesturing-and-advanced-biometrics.html]http://www.patentlyapple.com/patently-apple/2015/07/intel-micron-introduce-3d-xpoint-memory-that-will-advance-8k-gaming-siri-hand-gesturing-and-advanced-biometrics.html[/URL]
IDK, if this signifies anything ... The Keynote exec from Intel is named: Rob Crooke :rolleyes: