Intel and Micron Produce Breakthrough (non-volatile) Memory Technology

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Winston

Lorenzo von Matterhorn
Joined
Jan 31, 2009
Messages
9,560
Reaction score
1,748
SANTA CLARA, Calif., and BOISE, Idaho, July 28, 2015 – Intel Corporation and Micron Technology, Inc. today unveiled 3D XPoint™ technology, a non-volatile memory that has the potential to revolutionize any device, application or service that benefits from fast access to large sets of data. Now in production, 3D XPoint technology is a major breakthrough in memory process technology and the first new memory category since the introduction of NAND flash in 1989.

The technology is up to 1,000 times faster and has up to 1,000 times greater endurance than NAND, and is 10 times denser than conventional memory.

3D XPoint technology will sample later this year with select customers, and Intel and Micron are developing individual products based on the technology.


https://newsroom.intel.com/communit...micron-produce-breakthrough-memory-technology
https://newsroom.intel.com/docs/DOC-6713
 
I just wonder if they are going to make them here or Malaysia. Overseas the security problems are much greater than if they're made here. I hope they make them here, we need the jobs....
 
I just wonder if they are going to make them here or Malaysia. Overseas the security problems are much greater than if they're made here. I hope they make them here, we need the jobs....
They'll be making them in their Utah facility. From the video conference cost per bit will be somewhere between NAND flash and DRAM.

3dxpoint-100598796-orig.png
 
Their production facility is in Lehi, Utah which, interestingly, is also the location of that new, huge, controversial NSA data center. How convenient, a huge customer right next door:

Micron facility:

https://www.heraldextra.com/news/lo...cle_7600f1e4-f532-50a6-b9c2-d6103371a44a.html

NSA Data Center:

https://fox13now.com/2013/06/06/nsa-data-collection-turns-attention-to-utah-data-center/

During the press conference it was asked if this innovation would negatively affect their NAND flash sales and they answered that it wouldn't. If the price per bit is as they claim to be less than NAND flash then I simply cannot see how that is possible. Even it was the same price per bit, 1000 times faster, 1000 times more write cycles, would add up to must-buy thumb drives and SSDs.
 
And to think I invested in bubble memory. That's really okay because I then sold my Apple stock for 7$ a share. Now all my money is in cardboard tubes and AP. A sure investment this time!
 
[video=youtube;ARJ8cAGm6JE]https://www.youtube.com/watch?v=ARJ8cAGm6JE[/video]
 
Yes, I saw that, too, this morning. Perhaps, Y.C. Chen, C.T. Chen, J.Y. Yu, C.Y. Lee, C.F. Chen, S.L. Lung and Rich Liu of Macronix International Co. Ltd., Hsinchu, Taiwan had an idea that Intel/Micron perfected and brought to market?


....Micron did get it's start swiping Texas Instrument's templates and technology....just saying.....
 
....Micron did get it's start swiping Texas Instrument's templates and technology....just saying.....
Not true. TI divested themselves from the commercial memory business and sold the unit, lock, stock and barrel to Micron Technologies. A former co-worked went to TI to develop the 64 MB memory chip, and once it was developed, he and his fellow TI employees had 2 options: 1.) move to Idaho and work for Micron, or find another job! He went to Micron and never looked back and he's still there 2+ decades later.

TI also divested themselves of their IR sensor business about the same time, and sold the entire unit to Raytheon.......

Bob
 
Not true. TI divested themselves from the commercial memory business and sold the unit, lock, stock and barrel to Micron Technologies. A former co-worked went to TI to develop the 64 MB memory chip, and once it was developed, he and his fellow TI employees had 2 options: 1.) move to Idaho and work for Micron, or find another job! He went to Micron and never looked back and he's still there 2+ decades later.

TI also divested themselves of their IR sensor business about the same time, and sold the entire unit to Raytheon.......

Bob


...and I don't live in Boise and remember the lawsuit they settled...at all...


I can also show you the basement on N.16th where the 6 principals met and the patent lawyer said, "Don't worry, it will take 20 years before we have to pay any lawsuit."
 
Last edited:
During the press conference it was asked if this innovation would negatively affect their NAND flash sales and they answered that it wouldn't. If the price per bit is as they claim to be less than NAND flash then I simply cannot see how that is possible. Even it was the same price per bit, 1000 times faster, 1000 times more write cycles, would add up to must-buy thumb drives and SSDs.

If the price per bit is between DRAM and Flash (as you state in a previous post) then this technology will be more expensive than flash and cheaper than DRAM (but still not as fast - from the articles I've read).

You can speculate then that you'll have very cheap bulk storage (flash), more expensive bulk storage (new tech), most expensive fast volatile memory (DRAM)
 
Kind of interesting, but still way too laggy. Need working NVRAM together with these SSD improvements.

Caches, caches all the way down to spinning rust until either the $/GB of SSD gets sub-2x that of HDD or else the rest of the system completely outpaces the SATA/SAS interface. Still tape for archives, though.
 
Interesting to see the quote from "2001 A Space Odyssey". Just rewatched that for the first time since the 70's when I was a teenager. I didn't appreciate it at the time but it still is a masterful movie, and considering when it was filmed it really was a harbinger of what was to come (some of it anyway).

If you haven't seen it, please give it a go (or a second go as in my case).

BTW, the flat panel displays (like our current LCD or OLED) were made using back-projection from a movie film. :)
 
but still way too laggy

Intel sees this as a replacement for BOTH the SSD and DRAM - so not fast enough to replace the cache, but plenty of speed and low enough latency for everything else.
Density and cost SHOULD be in line with these usages too....but that all depends on yield and where marketing decides to price things.
 
but still way too laggy

Intel sees this as a replacement for BOTH the SSD and DRAM - so not fast enough to replace the cache, but plenty of speed and low enough latency for everything else.
Density and cost SHOULD be in line with these usages too....but that all depends on yield and where marketing decides to price things.
Intel -marketing- sees this as a replacement for the RAM, but....
DDR4 quad channel RAM: 50,000MB/s, nanoseconds
Optane SSD: 2,000MB/s, 100μs
 
Intel -marketing- sees this as a replacement for the RAM, but....
DDR4 quad channel RAM: 50,000MB/s, nanoseconds
Optane SSD: 2,000MB/s, 100μs


Be careful not to mix raw technology capability with the limitations of the "off the shelf" controller IC used in the first implementation.
New chipsets are coming with raw XPoint controllers inside.
 
Be careful not to mix raw technology capability with the limitations of the "off the shelf" controller IC used in the first implementation.
New chipsets are coming with raw XPoint controllers inside.

Definitely. Don't get me wrong, I'm almost as excited by this as I am by the HP stuff. It fills a nice gap in the caching hierarchy, and that's awesome!

It's just that the shipping Optane product varies from the raw specsheet by orders of magnitude ; and the raw XPoint specsheet varies from raw RAM by at least an order of magnitude. As long as there's a PCIe bus in the way, we're talking hundreds-of-micros not tens-of-nanos.

Do I hope to see XPoint-like sticks in RAM-like slots in the next 3-5 years? Yes, yes I very much do.
 
Back
Top