Researchers: We Make Transistors 1M Times More Efficient

[Max Collodi]

[citation][nom]nuvon[/nom]Please do NOT patent this before you actually physically make it![/citation]
McCormick already owns the patent on Spin-Transistors and has successfully used them to perform the same operations as a conventional CMOS circuit.

 

[derekullo]

[citation][nom]GeoMan[/nom]I fail, that's what I get for trying to think before my first cup of coffee (-;[/citation]

"It seems as though my state-funded math degree has failed me."

 

[memadmax]

10-15 years sounds about right.
You have to finish out the research, and after rounding out the legal crap, send it off to a company. The company in turns figures out how to put it to work, starting off with known principles/architecture/design etc at first. Next they have to to build all new fabrication machines to build the first chips. Then after a year or so start rolling a production run and go from there...

 

[dalethepcman]

This would be totally awesome, too bad the worlds going to end in 5 months, so we will never see it, lol.

 

[blazorthon]

[citation][nom]nuvon[/nom]Please do NOT patent this before you actually physically make it![/citation]

What's wrong with patenting your own ideas, especially ones that are not obvious? Now that they've announced them, if it's not patented or such in some way, a better funded project could be started by someone else to copy and get the technology to market before the inventors of it do and then keep them out of the market by stealing the concept. There' nothing wrong with patenting something like this be it on paper or not at this time. Now if they don't go anywhere with it, then there can be problems. However, why shouldn't they be allowed to protect their ideas as they're working on them? That would be unfair and would go against the intended spirit of the patent system rather than the perversion of it that we see most companies using it as.

 

[serendipiti]

Sorry, but do you realise how do they manage to set the spin of an electron (Scanning Tunneling Microscopes) ? and do you think that once you show you can use it as CMOS, that means you can build circuits made of billion spin transistors configured with complex patterns ?, and set reliably one and only one spin there ?
Sorry, but this is not in a 15 year timeframe...

[citation][nom]blazorthon[/nom]What's wrong with patenting your own ideas, especially ones that are not obvious? Now that they've announced them, if it's not patented or such in some way, a better funded project could be started by someone else to copy and get the technology to market before the inventors of it do and then keep them out of the market by stealing the concept. There' nothing wrong with patenting something like this be it on paper or not at this time. Now if they don't go anywhere with it, then there can be problems. However, why shouldn't they be allowed to protect their ideas as they're working on them? That would be unfair and would go against the intended spirit of the patent system rather than the perversion of it that we see most companies using it as.[/citation]

because you can't patent that idea. I can patent a new mounting / construction way for doors, but I can't patent the fact that wood is useful for making doors...

 

[blazorthon]

[citation][nom]serendipiti[/nom]Sorry, but do you realise how do they manage to set the spin of an electron (Scanning Tunneling Microscopes) ? and do you think that once you show you can use it as CMOS, that means you can build circuits made of billion spin transistors configured with complex patterns ?, and set reliably one and only one spin there ? Sorry, but this is not in a 15 year timeframe... because you can't patent that idea. I can patent a new mounting / construction way for doors, but I can't patent the fact that wood is useful for making doors...[/citation]

Your argument is like saying that they're patenting transistors themselves. They're not patenting transistors, they're patenting a specific type of transistor. It's kinda like patenting a specific architecture for the transistor design. There's nothing wrong with that.

 

[PreferLinux]

[citation][nom]coolestcarl[/nom]BTW, "Moore's law" is not really a law in scientific terms, its was just Moore's opinion that has been popularized as a scientific law. He was just an Intel engineer who just believed that transistor counts would double ever year based on historical observation at the time (1965).Its just been used as a clever tool to showcase Intel's development prowess in keeping up / exceeding expectations of the "law".Still quite nice to know that we are nearly on the brink of unleashing awesomely powerful digital overlords on the world. All hail our cold calculating masters! We bow to your superior logic![/citation]
I thought he was one of the co-founders?

 

[Guest]

If we follow Moor's observation, 2 by the power of 20 makes roughly a million. So in 20-40 years maybe, if we manage to convert electron spinning into electrical or optical signals and vice versa. Otherwise it will be as good as a drunk cowboy and don't spin me around too fast, as I can fly off the dance floor )))

 

[master_chen]

OPEN THE BAY DOORS HAAAL!!!

 

[stevo777]

Spintronics research has been going on for years--at least a decade--it is not new. There are tons of obstacles to overcome in the research phase, and, even when they are overcome, there will be a ton of obstacles to overcome to get it to market. Still, cool research.

 

[optimuscream]

I can imagine that .. playing Battlefield 5 in 42 inch monitor connected to cigarette box size computer with all graphic setting in ULTRA QUALITY ..