A Nanowire Memory Cells Has Been Discovered For Superconducting Computer. This is a great leap in computer technology.
http://www.sciencetimes.com/articles/17153/20170616/nanowire-memory-cell...
Subscription Note:
Choosing to subscribe to this topic will automatically register you for email notifications for comments and updates on this thread.
Email notifications will be sent out daily by default unless specified otherwise on your account which you can edit by going to your userpage here and clicking on the subscriptions tab.
If that truly scales up to mass production, then we might finally start seeing real advancement in the top end power computing platform again. (Advancement in high end, desktop/server grade CPU has mostly stalled in the last 5 years. Where they just add more cores and efficiency instead of raw power.)
Greatly reduced heat dissipation is actually huge, we been stuck at around 4 ghz for quite a while now. If this allows for cores to suddenly run at 5 Terra-hertz or faster, we will see a sudden large leap in processing power. Also low heat dissipation probably means much less power used, meaning laptops, cellphones, etc can run much longer on current battery technology, or run much more powerfully/and or both.
An advancement like that could also allow for mobile, goggle sized, (and smaller) VR/AR devices that are un-tethered that were at a reasonable price point, which would power the next electronic revolution, like smartphones did. With as big, or bigger cultural impact.
Want to hear something scary? Think about this:
Lets say a core is 2 cm wide. Electrical signals travel at 3*10^10 cm/s. So in a second it can travel across the core (assuming straight line) 1.5*10^10 times. Which means that core can preform at most 1.5*10^10 cycles or 15 gigahertz. So we are within an order of magnitude of that already.
Yikes on that 15 ghz speed limit, perhaps nanowires allow for higher electric signal speed then through silicon based processors?
Of course they can design processors that the electrical signal rarely has to travel across the entire chip for the chip to perform an operation. But admittedly my knowledge on this stuff is a bit limited.
Faster than the speed of light? This is kind of the point; it don't matter what you make it out of, we know that speed is the upper limit.
--------------------
Perhaps but that would require fundamental changes, since most operations are dependent on the outcome of the previous operation (so you have to be sure it had time to complete). That is why adding a second core (as you mentioned) doesn't double your computing power; when operations depend on the previous one, you don't get any performance by adding cores. Consider the problem tan(cos(sin(π/2))). A million cores won't speed up the calculation of that problem since every calculation depends on the outcome of the previous.
--------------------
I think the place where improvement could be gained is by making them smaller. Make it 1/10 as wide and you could get 150ghz (as a maximum upper limit anyway).
Graphene transistor could mean computers that are 1,000 times faster
Next-gen, carbon-based transistors would far outperform today's silicon ones
https://www.sciencedaily.com/releases/2017/06/170613145144.htm
Ah, you mentioned electrical signals. I could be wrong, but, electrical signal moves a lot slower through silicon based chips then light moves in a vacuum?
Also read about them designing 3d layered chips, instead of everything on one plane, that would reduce travel time, but up complexity greatly. With them being stuck at 4-5 ghz for so long they had no reason to design around speed limitations in that way.
It is confusing. Let me make an analogy with sports fans doing the wave at a stadium. When the person next to you does the wave, then you do the wave say 1 second later. If the fans are 2 meters apart then the wave propagates around the stadium at 2 m/s while the fans don't really do much movement individually. When the electrons in a wire begin moving (at a relatively slow speed), the rate at which this movement propagates along the wire (propagates to the next electron) is essentially the speed of light. In practice it moves something like 60% to 99% the speed of light; so when I put in a value to calculate the maximum ghz I used the speed of light (meaning the material of our fictional sci-fi core is perfect). This is why I said it is the best you can do.
oop double post
This is the next wave in computer tech. Soon will have advanced computer AI's.
Nanowires, the future of electronics
https://www.sciencedaily.com/releases/2017/06/170607123946.htm
science is cool