Researchers Achieve Amazing Memory Density

Mr. Fahrenheit writes in with a Wired story on research out of Arizona State, where researchers have "developed a low-cost, low-power computer memory that could put terabyte-sized thumb drives in consumers' pockets within a few years... The new memory technology — programmable metallization cell (PMC) — comes as current storage technologies are starting to reach their physical limits." PMC involves the on-demand creation of copper nano-wire bridges. It's said to promise memories that are 1/10 the cost and 1/1000 the power consumption of conventional Flash memory. Three memory manufacturers have licensed the technology and the first chips are expected on the market in 18 months.

Comments

And it will be released in 5 years

[Daimanta]

Togheher with your flying car. No. Really.

Vaporware.

[The Iso]

We've all seen this a dozen times before. All "amazing density storage" is vaporware, even if we'll be able to buy it real soon now.

Re:Vaporware.

[rbarreira]

I don't think most people nowdays appreciate how much 1.44 MB is...

Almost Infiniate?

[WillRobinson]

"Kozicki says the process is like condensing a crystal from a solution, except that the process is almost infinitely reversible. If the PMC is fed a positive charge, the copper atoms return to their previous free-floating state, and the nanowires disassemble."

I would like to know the exact number of cycles this will take, plus or minus a few million times.

The technology looks like it would eventual deplete the material used for the interconnect. But than again I am not a physicist.

Re:Other specs?

[Enlightenment]

Whoops. That should read:

Kozicki says the process is like condensing a crystal from a solution, except that the process is almost infinitely reversible. If the PMC is fed a positive charge, the copper atoms return to their previous free-floating state, and the nanowires disassemble.

From TFA. Wouldn't this imply that they think its mean time to failure is pretty long? Of course, they didn't say anything about speed or durability. But nanoscale changes should happen pretty fast, right?

Cost vs. Price

[Boogaroo]

It may cost 1/10th the cost to make, but I submit that we'll be charged double the current price simply because it's "new and improved." Just look at CDs vs. Tape or VHS vs. DVD.

Re:Cost vs. Price

[Artraze]

I doubt it. CDs were a different product, and therefore a chance to extract mondo bucks from consumers. The "new and improved" argument had a little to do with it, but copyright and forced obsolescence of tapes was what actually allowed for the higher price. (i.e. Consumers had to buy the expensive new option because they had no choice otherwise, not because they were willing to pay the artificial premium for the newness.)

For this, however, there is no similar mechanism. To most consumers it will just look like a normal flash drive and work like a normal flash drive. Joe Sixpack doesn't care about the technology, and probably doesn't even know flash dives have limited write cycles (not that he'll ever approach them). Unless the new drives offer more memory or a better price, there will be no reason to buy one.

Of course, in the embedded market, this would be huge due to reduced power consumption and write cycles (which eliminates the need for wear leveling). Also, for more extreme environments (I'm looking at you, space) the fact that this memory changes physically and doesn't simply hold charge (which is rather easily changed) is also a major plus. Even with these advantages, I doubt that there will be any sort of price inflation in these markets either since these buyers know what they're doing.

Read/write (especially WRITE) speed?

[SamP2]

Cheap? Cool. Large size? even better. Energy efficient? Meh, I'm not in Greenpeace, but sure. And I'm even willing to believe it's reasonably reliable.

But how come nobody's concerned aobut the the IO speed? I wouldn't be too concerned about reading, but if writing/rewriting requires real-time rebuilding of gates, wouldn't it be snail-slow?

The IO of even regular hard drives already becomes a significant factors as drives grow exponentially larger and speed stays the same as always. If this is even slower, it'd become a serious deterrent.

A politically incorrect question

[saltydog56]

The technology sounds great, and if they come through with it I am sure it will lead to many innovations. However, am I the only one who feels a little uncomfortable with research done at a state university, funded by the public, and performed by unpaid or low-paid grad students being licensed by "Arizona States business spin off, Axon Technologies"

I know that type of arrangement may be common place today but I sure would like to follow the money trail.

Energy efficiency not meh

[EmbeddedJanitor]

Energy efficiency is not at all arbitrary if it is coming out of a battery.

Re:Other specs?

[NemoinSpace]

How about letting them build the thing first? Or do you suggest we form a statistical opinion based on the two or three prototypes that might exist? #places in circular file under vaporware for 18 months.

Re:Mac Time Machine - rsync for dummies

[Solra Bizna]

rsync makes incremental backups?

-:sigma.SB

Re:Sorry to be a spoil-sport, but

[Waffle Iron]

If I'm not mistaken, the signaling delay of conventional circuits is dominated by the reactance of the electromagnetic fields, not by the momentum of the electrons. Therefore, there's not much basis to conclude that the momentum of copper atoms moving over a couple of nanometers distance will cause a significant delay reletave to an electronic circuit saddled by its capacitance and inductance.

Ummm, why?

[phorm]

When would you ever have to transfer a full terabyte at a time? Unless you're doing a really bigass backup to this thing, you probably won't.
And if you are, well that's a hell of a lot faster and more convenient than burning 233 standard DVD-R's (about what it would take with non dual-sided discs) or writing the equivilent tape or network-based backup method. Heck, that beats out most disk-to-disk transfers.

New Game Delivery System

[Kingrames]

Since this probably means that game producers will be able to put their games on flash drives instead of CDs and DVDs, it would be even more convenient than having a backup disk.

That, and they'd be able to shrink down the size of game boxes again, from dvd size to, dare I say it, cigarette pack sized. Your next video game could be dispensed by a vending machine.

Re:Other specs?

[Dr_Barnowl]

could build a RAID and replace hard drives every few years and still come out ahead price-wise. Everyone repeat after me ;

"RAID is not a replacement for backup."
"RAID is not a replacement for backup"

RAID does not protect you against rm -rf / , or another idiotware/malware.

Re:Other specs?

[PopeRatzo]

How about speed, durability, mean time before failure, etc

Man, you guys are a tough crowd. This is a breakthrough for chrissake. I can imagine if Slashdot had been around when they reported Alexander Graham Bell's famous "Watson, come here I need you" experiment. You'd have been saying "But will he be able to get speech enhancement using minimum mean-square error log-spectral amplitude estimators?" And asking about Wiener filters.

But that's why I love you.

[he said "Wiener" filter, heh-heh]

Re:Other specs?

[Machtyn]

I hope you realize that wasn't his point. His point was, he could build a RAID setup to get 1.6TB, replace the drives once a year and still come out ahead when the 1.6TB single drive comes out. RAID 1 is not the only RAID out there.

I also get your point. RAID 1 is fault tolerance, not backup solution.