Slashdot Log In
Nanotube Memory Finally Beats Flash For Speed
Posted by
ScuttleMonkey
on Fri Feb 06, 2009 02:50 PM
from the more-nano-for-your-nano dept.
from the more-nano-for-your-nano dept.
holy_calamity writes "Although flash memory that stores each bit on a single nanotube has been tinkered with in the lab for years, it has always been much slower than the devices in use today. A Finnish team has now cracked that, demonstrating single bits of nanotube memory that can be written in just 100 nanoseconds. Existing flash memory takes tens of microseconds."
Related Stories
[+]
Technology: Carbon Nanotube-Based NVRAM In 2-3 Years? 66 comments
According to NanoWerk, UC Riverside researchers have come up with a memory device based on telescoping multi-walled carbon nanotubes. According to one of the researchers, 'This finding leads to a promising potential to build ultrafast high-density nonvolatile memory, up to 100 gigahertz or into the terahertz range" and a prototype could be demonstrated "in the next two to three years.' Similar devices from UCLA and Caltech based on bistable rotaxanes are farther along in being integrated into actual memory circuits, but tend to break after a fairly small number of position changes. Carbon nanotubes may promise more durable switches.
[+]
Technology: New Nanotech Fabric Never Gets Wet 231 comments
holy_calamity writes "New Scientist reports on a simple coating for polyester that renders it unwettable — even after two months underwater it emerges dry to the touch. Water cannot attach to the new fabric thanks to nanostructured filaments and a structure that traps a constant air layer. One potential use is for low-drag swim wear."
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
And what exactly (Score:4, Funny)
Re:And what exactly (Score:5, Funny)
Parent
Re: (Score:2)
Re: (Score:2)
Computers are SO unneccessary, yes ....
So how did you post this ? Chisel it out on a block of granite, and Fedex it to slashdot headquarters ?
What a whackjob !
PETA won't like this at all.... (Score:3, Interesting)
I can already see the craigslist ads..."Wanted: Computer geek to come snake out my RAM."
Re: (Score:3, Funny)
That's what she said!
Re: (Score:3, Funny)
what will we do when these "tubes" become clogged, and we can't get our internets?
Fixed it for you, Convict Stevens.
Re: (Score:1)
(for the less cultured of the
Re: (Score:3, Funny)
Re: (Score:2)
Great, just another set of tubes to tie. Like I'm made of money.
Wow, that's pretty cool (Score:5, Informative)
Call me back when it's available at Costco for 100$ per Terabyte.
Re: (Score:3, Insightful)
Heck call me when a product is availible....at ANY price
Re:Wow, that's pretty cool (Score:5, Funny)
Heck call me when a product is availible....at ANY price
Sure. You can order it now for the low low price of $3 billion. Shipping time is estimated at 10 years and 7 days with free standard shipping, or 10 years and 3 days with Express Shipping for $15 extra.
Parent
Re: (Score:2)
COOL!
Who do I make the check out to? BTW it is post dated Feb6, 2019.
Re: (Score:2)
Make it out to me. I don't mind waiting until I'm 36 to become a billionaire.
Re:Wow, that's pretty cool (Score:4, Informative)
I RTFA yesterday, it will be a while. Right now they only have a one bit memory, and TFA says a lot more work will have to be done before they can get millions or billions of them on a chip, let alone mass produce the thing.
Parent
Re: (Score:1, Informative)
You read /. but don't care about new hardware until its something you can buy for cheap? Okay... There are some hobbies you might consider, like stamp collecting, to more effectively use your time.
Re: (Score:2, Funny)
That's a very ineffective use of time. It can take minutes just to make reliable oral contact with a cock that is in the process of fucking, and once achieved this contact is rarely any improvement on the previous status of the cock even in those cases where it is not an uncomfortable hindrance.
This is, of course, assuming that the orifice in question cannot accommodate a human head positioned such that sucking is possible from the inside. In that case, it's really questionable if what's happening can be ca
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
That should... (Score:5, Funny)
Re: (Score:2)
I think I can remember a day when system memory was measured in the 100s of nanoseconds.
Luxury! (Score:4, Interesting)
100s of nanoseconds? Luxury! When I were young you had to store your data as sound pulses in a tube full o' mercury. You had to wait 100s of milliseconds for the pulse to reach the other end. Tell that to kids nowadays, they won't believe you.
Parent
Does that mean... (Score:1)
... that Ted Stevens is a prophet and the internet WILL be a series of tubes?
Nano-Tubes?!?!?! (Score:1, Funny)
Is this like some sort of scaled down version of the internet?
nano internet? (Score:2)
Re: (Score:2)
Re:nano internet? (Score:5, Funny)
No, it would be 10^-9 LoCs. You're thinking of deci-internets.
Parent
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Hmmm. I'm Asperger only on days with at least two vowels in the name.
Re: (Score:2)
That's libraries [angryflower.com].
"The library's open today"
"The libraries are all open"
"The library's books are all open"
"All the libraries' patrons the world over are open"
Low OPs lifetime (Score:4, Insightful)
"The device managed to withstand 18,000 operations, which is a reasonable lifetime for a memory device, she adds."
Is that good for experimental chips or do I not understand how such a low number is reasonable?
Re: (Score:2)
18,000 operations should be more than enough to fill this thing full of porn, there's no need to get greedy.
Re:Low OPs lifetime (Score:5, Informative)
Your average Flash chip does 100k erase/write cycles. 18k is certainly reasonable for new tech, which will certainly improve over time. The number refers to the number of operations per erasable block (or it will in the future), so in practice you get a much larger number of total I/O operations on the entire chip, given a reasonable wear leveling algorithm.
Parent
Not there yet... (Score:3, Interesting)
How much latency does that algorithm add? They are only testing one bit. Won't a controller and the wear-reducing slow it down a fair amount?
Still a ways to go:
"The next challenge is to join an array together into a working memory chip, as the team has so far only tested single carbon nanotube elements. And although they have only proved capable of "remembering" data for several days after the power is cut, the team are confident this can be extended."
Several days is a pretty short life for SSD...and longer
Re: (Score:3, Informative)
Let's say they make this into a thumb drive. Now, let's say that you read/write the entire drive twice a day. That's four operations. 18,000/4/365 gives you twelve years of this. Even if you are filling and then erasing the drive ten times every day it is still two and a half years of use. Less than you can do with current flash memory, but fine as a proof-of-concept.
Re: (Score:2)
CD-RW will only last for 100 writes. Have you ever seen a failed CD-RW? I didn't. Even if you had a nanotube drive that you completely overwrote every hour, it would last for 2 years. An USB drive with 100k writes would last for 11 years. In real life scenarios drives are rarely written to more than a few times a day, and certainly not in the same places. This can be further reduced with wear leveling filesystems. The only reason we don't use those widely is that it's extremely hard to write a filesystem dr
Re: (Score:2)
CD-RW will only last for 100 writes. Have you ever seen a failed CD-RW? I didn't.
I used packet writing quite a bit for a while and I think I killed a few CDRWs with it over the years when I used it.
However the vendors fucked up on thier software. Adaptec and ahead came up with incompatible packet writing software and even if they had been compatible bundled copies of nero didn't generally come with incd. So as the burner vendors switched to bundling nero packet writing died out at least among those of us to
Well, which Flash are we talking about here? (Score:4, Funny)
Jay Garrick isn't as fast as Wally West or Barry Allen...
Regressing (Score:5, Funny)
ENIAC was orders of magnitude faster. These guys do 100 nanoseconds/nanotube -> 100 seconds/tube!
Nantero... (Score:3, Interesting)
What ever happened to Nantero? Weren't working on this a few years ago?
So, how long to build a fab plant? (Score:1)
Say for about 1 million units a year, with continuous production ...
What, you can't get the loan for a fab plant for this non-production memory?
And it will take 4-5 years to build it?
I'm thinking we better make the plant design fusion powered ... cause it's going to be a while, and we might as well be pie in the sky about it.
Re: (Score:2, Informative)
Existing flash memory takes tens of microseconds.
Re: (Score:3, Insightful)
Where are all the fuckers who bitched about me when I said most people can't understand their beloved SI units for shit (let alone read in this case)?
Re: (Score:2, Funny)
Re: (Score:2)
Re: (Score:2)
Is that an imperial shit ton, or a metric shit tonne?
Re: (Score:2)
Since "digital" usually is an interpretation of analog, I'd get used to seeing both of them around for a long time.