backtop

Print E-mail del.icio.us 16 comment(s) - last by Vinas.. on Nov 3 at 12:54 PM

One layer of the PCMS array. The memory cell stack, including rows and columns, is shown sandwiched between M2 and M3  (Source: Intel)
Will take a few years to bring to market, though

Intel is one of the market leaders in SSDs thanks to its supply of 34nm NAND flash memory from IM Flash Tech. The world's largest semiconductor company owns half of IMFT in partnership with Micron Technologies. IMFT's NAND flash chips are not only used in SSDs, but also in USB flash drives and embedded devices like cell phones.

The price of NAND flash memory chips has been dropping every year, as bit densities increase and new process technologies are introduced. However, the NAND flash party is expected to end in a couple of years, as there is a limit to how small flash chips can scale.

It is expected than NAND flash will not scale beyond the 22nm node. Even at 3xnm, there are only a few electrons available for the Fowler-Nordheim tunnel injection process, which is how writes are performed on NAND flash. Current SSDs are already running at 20V internally because of this. SSD companies are already having speed problems with Samsung's 32nm NAND flash chips, causing a delay in their market adoption.

Intel has been thinking about this problem for a while, and has been working with Numonyx on phase change memory. PCM uses chalcogenide glass, which can be altered between crystalline and amorphous states with the application of heat. Intel believes than PCM will be the successor to NAND flash, with greater capacity, lower power consumption, and higher speeds.

“We continue to develop the technology pipeline for memories in order to advance the computing platform,” said Al Fazio, Intel Fellow and Director of Memory Technology Development.

Recent research has yielded a stackable 64Mb test chip. The ability to stack multiple layers of PCM arrays within a single die is something that will bring higher densities. Intel is demonstrating a vertically integrated memory cell which it calls PCMS (Phase Change Memory and Switch). PCMS is comprised of a PCM element layered with an Ovonic Threshold Switch (OTS) in a cross point array. A cross point array is formed by switch cells occupying separate blocks in a common plane of an integrated circuit.

“We are encouraged by this research milestone and see future memory technologies, such as PCMS, as critical for extending the role of memory in computing solutions and in expanding the capabilities for performance and memory scaling," said Fazio.

Intel says that it has built PCM chips with low latencies and high bandwidth PCM, and found that it can be stable down to 5 nanometers. It will still be a couple of years before this is refined enough to go to market, but Intel wants to be ready to pick up the pieces when NAND flash scaling stops.

Additional information about the memory cell, cross point array, Intel's experiments and results will be published in a joint paper titled “A Stackable Cross Point Phase Change Memory”. It will be presented at the 2009 International Electron Devices Meeting in Baltimore, Maryland in December.

Comments     Threshold


How about light?
By SiliconJon on 10/29/2009 12:53:05 PM , Rating: 2
How about instead of heat, and even instead of electricity, we seek a material that behaves similarly but is resonant to light for its phase change initiation. Heat just sounds like a bad idea for the control mechanism of such a device to me. Light would also be much more durable, and that's where it's all going, eh?




RE: How about light?
By omglol on 10/29/09, Rating: -1
RE: How about light?
By DrizztVD on 10/29/2009 3:21:52 PM , Rating: 4
quote:
Don't believe me? Ask your friendly neighborhood physical chemist. Even an organic chemist should know that a reaction requiring "hv" needs heat or light.


You've got it wrong. You are explaining the mechanism of heat transfer through space, which involves EM waves. But saying heat and light is the same thing is incorrect. The temperature of an object is determined by the momemtum of it's particles. Exposing it to light can increase it's temperature- but that's after light energy has been converted into heat-energy .

The PCM cell works by changing its electrical resistivity. This can only be done by varying its temperature - which changes the physical ordering of the material's particles. You can use light to produce the heating - but that is a very poor choice in this case.

What the above reader means to say is that the PCM should function by measuring transparency (opacity) of the PCM cell rather than resistivity. This can be done in theory - but the current technology still has too many advantages to what the emerging photonic industry can do.


RE: How about light?
By omglol on 10/31/2009 7:32:09 AM , Rating: 1
quote:
mechanism of heat transfer through space

You correct me with this?
Heat is the transfer of thermal energy itself in any of its forms.
A quick wikipedia search confirms: "...heat is the process of energy transfer from one body..."
quote:
But saying heat and light is the same thing is incorrect

No. Saying they're always the same thing or saying that heat is only light or light is only heat would be wrong. What I said was simply incomplete, not completely incorrect. Note that you, too, can explain things incompletely. You say that IR is the transfer of heat through space. Not only is that wrong as I pointed out above, but it's incomplete because you're not including the fact that IR can transfer thermal energy through any media that is transparent to it.
quote:
temperature of an object is determined by the (sic) momemtum of it's particles

Actually it's the average kinetic energy of its molecules, but you're close.
Also, "its" and "it's". Look them up.
quote:
Exposing it to light can increase it's temperature- but that's after light energy has been converted into heat-energy

You mean "its" and "thermal energy".
quote:
which changes the physical ordering of the material's particles

Resistance is related to temperature? Who would have known?
quote:
above reader means to say is that the PCM should function by measuring transparency (opacity)

False. I wasn't saying this device measures the %T or the ABS of the material in IR at all, I was saying IR radiation is a product of molecules with high thermal energies.

You need to read better. I wasn't attempting to describe the function of this device. All this because I wasn't trying to be particularly anal retentive about my explanation - and that's why the people who post here are pricks, no person of any value or with anything valuable to accomplish could spend this much time explaining these simple concepts to children on the internet.


RE: How about light?
By Vinas on 11/3/2009 12:54:51 PM , Rating: 2
You noobs might want to Google William Herschel to read about his experiment with a prism and the discovery of infrared light/heat.


RE: How about light?
By BigPeen on 10/29/2009 3:23:30 PM , Rating: 2
That is the worst description of "heat" I've ever heard.
E-M radiation comes in quanta known as photons. "Heat" can come in several different forms, namely, any form of energy transfer (photons, phonons, etc.)


RE: How about light?
By omglol on 10/31/2009 7:40:56 AM , Rating: 2
You've never had any formal education then? I'm sorry.

I was only mentioning the IR transfer of q. Clearly I could have created a multi-page explanation on every form of energy transfer known, but since I was already on a tangent, I kept it short and sweet.


RE: How about light?
By omglol on 10/31/2009 6:24:11 AM , Rating: 1
lol, I guess if you want to be "taught" something, you can also fail to fully explain it on a forum of self-important nerds when you're trying to make a complex concept into something shorter and understandable.
I guess you folks think that heat is transferred by only one method depending on the medium through which it is being transferred. That's true if you're talking about a perfect vacuum, otherwise there are always going to be several methods of transferring thermal energy, including infrared photons. Clearly there are other methods, but, since I didn't mention them, I'm wrong/evil/scolded/etc.
Too bad what I said in my post isn't even wrong, just not complete. So, sod off, self-righteous nerds.


At the picture
By PublixE on 10/29/2009 10:03:14 AM , Rating: 2
"One layer of the PCMS array. The memory cell stack, including rows and columns, is shown sandwiched between M2 and M3 (Source: Intel)"

Their process looks very mature. I don't see very many anomolies. Maybe they are further then they say they are!




RE: At the picture
By menting on 10/29/2009 12:22:28 PM , Rating: 3
it does look pretty good in the picture provided. It doesn't however, say what process node this was built on. I won't be surprised if it was built on a much older node for research purposes.


DailyTech Does It Again
By misbfa1 on 10/30/2009 10:00:08 AM , Rating: 1
Oh thank you, thank you, thank you, DT for reposting old news! I guess most people on here have not been reading DT as long as I have, or they would remember DTs OWN articles from three years ago, stating phase change memory was just a few years away.

From Samsung Sept 10 2006:

http://www.dailytech.com/Samsung+Announces+Phasech...

General PC article from Dec 11 2006:

http://www.dailytech.com/Phase+Change+Memory+Techn...

From Intel on March 9 2007:
In this article intel says PC mass production by the end of 2007.
http://www.dailytech.com/Intel+to+Sample+Phase+Cha...

From Intel Feb 7 2008:
Intel is supposedly shipping PC memory for testing

http://www.dailytech.com/article.aspx?newsid=16636

Do a search on DT for "phase change", and you will see that about once a year, for the past 3 years, DT posts an article on this to get us all riled up. Now, here we are Oct 29 2009, and PC tech is "just a few years to market". Typical DT, when reading this site you think that Star Trek level technology is always 5 years away.

Six technologies DT uses to get us excited on a semi-regular basis:

Phase Change Memory
Solar Power
Batteries/Capcitors
Holographic storage et al
Graphene
Carbon nano-tubes

Supercapacitors here in 5 years Aug 9 2006:
http://www.dailytech.com/New+Nanotube+Batteries+Re...

While researching the links for this post, I see that DT has a new article about Zinc-Air batteries with w/ 3x capacity here next year...SUPER LOL!




RE: DailyTech Does It Again
By really on 10/30/2009 2:17:18 PM , Rating: 3
What exactly is your point? Are they supposed to report on something once and then never again?


off topic
By semo on 10/29/2009 10:31:46 AM , Rating: 2
quote:
IMFT's NAND flash chips are not only used in SSDs, but also in USB flash drives and embedded devices like cell phones.
Anyone know who makes such USB flash drives?




Future of Computing
By XGene on 10/30/2009 8:48:00 AM , Rating: 2
So what you are saying is that basically our computers may eventually be running on some form of PMS.. Great!....




Technology
By Spivonious on 10/29/2009 9:33:56 AM , Rating: 1
Isn't it great? :)




The Whitehouse is not small!
By icanhascpu on 10/29/2009 9:55:59 AM , Rating: 1
Look at that picture, come on!




"Vista runs on Atom ... It's just no one uses it". -- Intel CEO Paul Otellini

DailyTech Poll
Which web browser do you use on your primary personal machine? 






44 Comments









botimage
Copyright 2009 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki