- Joined
- Oct 30, 2003
http://www.optonline.net/News/Article/Feeds?CID=type=xml&channel=32&article=13656314
Very interesting read.
Very interesting read.
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
HP Propose Alternative to Transistors
- Thread starter Avg
- Start date
- Joined
- Feb 25, 2004
- Location
- MA, USA
they say the have many more uses than normal transistors than don't say what. They say that they are better than transistors but don't say why other than they are smaller and are easier to fix defects with. But good way to think outside of the box. What temperatures do they run at is my questions and what does this mean to the overclocker. To me new technology means more expensive.
- Joined
- Sep 16, 2003
- Location
- Knoxville, TN
Well I suppose Moors law has some life left in it 
. This is good news indeed
. This is good news indeed
- Joined
- May 27, 2002
- Location
- Ohio
well at least its interesting. looks promising too. hope to hear more about this in the future.
and i'm sure my old Electronics teacher would love to hear about this.
and i'm sure my old Electronics teacher would love to hear about this.
- Joined
- Sep 16, 2003
- Location
- Knoxville, TN
Especially if the yeilds are betterAvg said:
- Joined
- Jul 30, 2002
- Location
- Midland, Ontario
Sentential said:Well I suppose Moors law has some life left in it
actually it means the death of Moors law since his law is based on the number of transistors not on actual speed
- Joined
- Jul 30, 2002
- Location
- Midland, Ontario
Looking at HP labs, this patent was filed in 2003, this news is just that they have done some theoretical experiments that actually show the crossbar will work someday
- Joined
- May 10, 2004
- Location
- Brooklyn
heh, yeah, there's always something better, and this may suceed transistors...
- Joined
- Jun 30, 2003
- Location
- Columbus, Ohio
Can anyone find any info on how crossbar latches are used to do logic? Is i like transistors where you construct logic gates?
- Joined
- Nov 25, 2001
- Location
- Niagara Falls, Ontario, Canada
Interesting.
As far as I understand things, you might only need 1-3 of these to perform a logic function that would typically be done by 6 or 8 transistors. I think they perform in some ways that are logically similar to electromechanical devices such as relays, which is what the earlier computers were built with, before they used vacuum tubes or transistors. The only drawbacks to electromechanical logic elements I think were speed and reliability. I believe this is similar to the technology that has been talked about as "molecular switching" At the simplest level I think each molecule acts just as a on/off 1/0 switch. You can see how this would apply to memory technology quite readily, that's all you need, a switch that can be set.
These also look like they will be very low power low heat devices. The density is amazing. If this were added to conventional CPU manufacture, then for an extra die area of only 2 square millimeters, one could get a memory array of something of the order of 80 terabytes
If this switching is static and long term, we could be saying goodbye to hard disks in 10 years. Although if an Xray laser diode were available it's possible that we could see densities within an order of magnitude, using similar molecular switching techniques, using a cheaper unorganised molecular deposition onto the disk, which would then be scanned and indexed optomechanically.
Actually thinking about it, it's conceivable that using an opto-molecular switching technology of even a quarter the density, one could make a 36 petabyte HDD (That's gotta be at least all the movies ever made all in HDTV format and MPAA think they've got probs now! ) with a miniaturised non-solid state laser, in the same form factor as a modern 3.5inch drive. And if the drive manufacturers are smart, they're doing that now and will have it out in 5 years (I'm thinking the data surface could be a 1cm x 6cm "roller" or drum at one end of the drive, and the laser could take up most of the rest of it, with opto-electrical scanning of the drum by the beam, either through liquid crystals that bend light or solid crystals that deform under electrical stress. ... )
hmmm, I'm checking my math again and again because that seems a so amazingly large amount of storage, but I keep getting the same figures for a 3nm device size. .. ... ....
Road Warrior
As far as I understand things, you might only need 1-3 of these to perform a logic function that would typically be done by 6 or 8 transistors. I think they perform in some ways that are logically similar to electromechanical devices such as relays, which is what the earlier computers were built with, before they used vacuum tubes or transistors. The only drawbacks to electromechanical logic elements I think were speed and reliability. I believe this is similar to the technology that has been talked about as "molecular switching" At the simplest level I think each molecule acts just as a on/off 1/0 switch. You can see how this would apply to memory technology quite readily, that's all you need, a switch that can be set.
These also look like they will be very low power low heat devices. The density is amazing. If this were added to conventional CPU manufacture, then for an extra die area of only 2 square millimeters, one could get a memory array of something of the order of 80 terabytes
If this switching is static and long term, we could be saying goodbye to hard disks in 10 years. Although if an Xray laser diode were available it's possible that we could see densities within an order of magnitude, using similar molecular switching techniques, using a cheaper unorganised molecular deposition onto the disk, which would then be scanned and indexed optomechanically.
Actually thinking about it, it's conceivable that using an opto-molecular switching technology of even a quarter the density, one could make a 36 petabyte HDD (That's gotta be at least all the movies ever made all in HDTV format and MPAA think they've got probs now!
) with a miniaturised non-solid state laser, in the same form factor as a modern 3.5inch drive. And if the drive manufacturers are smart, they're doing that now and will have it out in 5 years (I'm thinking the data surface could be a 1cm x 6cm "roller" or drum at one end of the drive, and the laser could take up most of the rest of it, with opto-electrical scanning of the drum by the beam, either through liquid crystals that bend light or solid crystals that deform under electrical stress. ... ) hmmm, I'm checking my math again and again because that seems a so amazingly large amount of storage, but I keep getting the same figures for a 3nm device size. .. ... ....
Road Warrior
