backtop


Print 42 comment(s) - last by mindless1.. on Jan 31 at 11:47 PM

Intel and Nanochip team up to develop 100 gigabytes per chip

This article was syndicated from Tiago Marques' blog at SiliconMadness.com

Nanochip, Inc., a Silicon Valley startup, has managed to raise $14 million in funding from Intel Capital, Intel's global investment organization, for further development of the MEMS technology.

You read it right: gigabytes, not gigabits.

According to Nanochip (PDF), the technology isn't lithography constrained, allowing production of chips of more than 1GB in capacity, in plants that have already been deemed outdated by current standards.

The lack of lithography constraints means cheaper products, resulting in an opportunity to also replace flash memory, as the technology is also non-volatile.

Today's factories should be able to produce the first products, estimated at 100GB per chip, when the technology is expected to be unleashed for public consumption by 2010. The first samples will be available during 2009.

PRAM, or phase change memory, was expected to be the technology to replace flash in the coming years, since it is also non-volatile, while it is much faster than flash. As Intel found out over the last few years, PRAM doesn't seemed to scale so well, in regards to density, and still has some boundaries to overcome -- namely it's thermal principles of operation.

Nanochip's details of the technology are ambiguous at best, though what is known is that the company is working on hybrid micro-electric-mechanical -- the partnership with Intel suggests a PRAM connection too.  The company has described this as a very small platter coated in chalcogenide is dragged beneath hundreds of thousands of electrical probes which can read and write the chalcogenide.  Casual estimates put this sort of density at one terabit per square inch, or 125GB per square inch. 

The company has not disclosed access speeds. That's a place where PRAM is appointed to be the undisputed king of the hill, so it could limit applications of this type of technology.

For now it seems that the flash SSD drives are going to be replaced before they even reach mass consumption -- which is a good thing. The technology is expensive, doesn't provide a lot of storage space and is prone to failure, due to the low amount of write cycles available per cell. Flash is perfect for pendrives and resisting shock, not so good for regular, intensive, HDD usage.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Conjecture
By Andypro on 1/28/2008 1:28:06 PM , Rating: 3
For now it seems that the flash SSD drives are going to be replaced before they even reach mass consumption -- which is a good thing. The technology is expensive, doesn't provide a lot of storage space and is prone to failure, due to the low amount of write cycles available per cell. Flash is perfect for pendrives and resisting shock, not so good for regular, intensive, HDD usage.

"Mass market" for a product scheduled 2 years away is more likely to hit in around 5 years, if at all. I believe SSDs will reach mass consumption long before then.

Flash drives are not at all "prone to failure," either. They use write leveling algorithms for even wear across the flash cells, and based loosely on the MTBF figures as well as life expectancy for SSDs, they still beat the hell out of any traditional hard disk drive. Meanwhile, this new mysterious tech, while interesting, has no hard data associated with its life expectancy or MTBF.




RE: Conjecture
By winterspan on 1/29/2008 6:17:21 AM , Rating: 2
I totally agree. What the hell is this writer smoking? He's obviously much too uninformed to be contributing articles....

To say that SSDs are "prone to failure" is just complete hogwash, especially compared to HDDs.
regular hard drives are *MUCH* more prone to failure than SSDs. SSDs in fact are VERY reliable. I wouldn't be surprised if the ratio of MTBF of SSD to consumer HDD is 10:1.

I'm sick of hearing this crap of flash memory having a limited read/write cycle. With wear-leveling technology, the SDD should on average easily outlast the piss poor reliability of harddrives.

On the topic of the actual article content, this new MEMS thing looks to be totally unproven in the real world, and just something that has been computer simulated. And yet the author claims that flash SSDs will be replaced before they are mainstream? HA!
2008 will be the year of upper-mainstream SSDs.
I foresee "regular enthusiasts" buying 128 SSDs by Q4.

anyways....


"So if you want to save the planet, feel free to drive your Hummer. Just avoid the drive thru line at McDonalds." -- Michael Asher

Related Articles













botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki