Why are we only now considering replacements for the transistor?

Orangey · 4 Jul 2011 at 19:38

http://www.physorg.com/news/2011-07-reveals-vital-insight-spintronics.html

Spintronics research began a long time ago. We knew CPU design would hit a wall so why have we left it so late to come up with a way forward?

As I get older I'm becoming increasingly bitter about the fact that I'm not seeing the future I was promised in science-fiction. We should have computers based off of crystals like in Stargate by now. And there aren't any flying cars either.

Dano · 4 Jul 2011 at 19:50

Yeh, I'm bitter they haven't cured baldness either

WingZero30 · 4 Jul 2011 at 19:53

Dano said:
Yeh, I'm bitter they haven't cured baldness either

You are bald

?

Gtiracer · 4 Jul 2011 at 19:55

WingZero30 said:
You are bald ?

Hence Homer Simpson as his avatar

WingZero30 · 4 Jul 2011 at 19:59

So Spintronics store data by spinning the electrons in an up-state or down-state and don't require external current (more flow of electrons) ?

Do physical spintronics exist?

WingZero30 · 4 Jul 2011 at 20:09

http://www.dw-world.de/dw/article/0,,15209635,00.html

"The challenge is now to build devices where the electron's spin is combined with the semiconductor technology", Heimbrodt told Deutsche Welle. "It may take another ten years at least before we find out how to combine the electron's spin with the semiconductor technology in electronic devices."

Seems like the aim is to combine the spin of electrons (spintronics) with the exisiting semi-conductor technology in order to build new electronic devices.

Still 10 years further. It must be something to with not able to spin electrons in solid state at high rate or something similar.

It seems to remind me of progression from HDD to SSD.

ScratchFive · 4 Jul 2011 at 20:21

http://www.youtube.com/watch?v=xPIbGq634yU

your right, the field has been around for a while, so I don't know how you can ask why we have left it so late? research, by its very nature, goes at its own pace...

Jokester · 4 Jul 2011 at 20:24

I would imagine just making existing silicon technology better has been so much cheaper than developing spintronics, but as you say we're now hitting a brick wall with traditional silicon semiconductors and the cost of developing new process nodes is growing massively with each reduction in manufacturing size so it makes development in alternatives more economically attractive and viable.

WingZero30 · 4 Jul 2011 at 20:33

Jokester said:
I would imagine just making existing silicon technology better has been so much cheaper than developing spintronics, but as you say we're now hitting a brick wall with traditional silicon semiconductors and the cost of developing new process nodes is growing massively with each reduction in manufacturing size so it makes development in alternatives more economically attractive and viable.

I think at the onset of spintronics, it is going to be very expensive as the field is indeed very challenging atm.

Secondly it will be interesting to see how Intel and AMD incorporate this technology in future cpus? Does this mean the end of multi-core era?

The emergence of spintronics may well lead to the emergence of new technology companies dedicated solely to spintronics manufacturing and could give Intel and AMD a hard time in future :cool:

Phenomenologica · 4 Jul 2011 at 21:43

The issue with articles like that is they have to dumb it down a vast amount for the public, as nobody who hasn't specifically studied it for a long time would be able to make heads or tails of the subject. Which means we get an impression of the concepts involved but not of the complexity of the task at hand. We're talking about storing information using one of the fundamental properties of an electron (Which incidentally isn't actually a physical spinning motion, it's just a metaphor). It has a lot of potential but the technical challenge of actually making it work is mind boggling. :eek:

namnoc · 4 Jul 2011 at 23:05

You should be upset they haven't come up with immortality. If they had you would have plenty of time to wait for everything else.

WingZero30 · 4 Jul 2011 at 23:12

Phenomenologica said:
The issue with articles like that is they have to dumb it down a vast amount for the public, as nobody who hasn't specifically studied it for a long time would be able to make heads or tails of the subject. Which means we get an impression of the concepts involved but not of the complexity of the task at hand. We're talking about storing information using one of the fundamental properties of an electron (Which incidentally isn't actually a physical spinning motion, it's just a metaphor). It has a lot of potential but the technical challenge of actually making it work is mind boggling.

Yeah I was also thinking along the same line. This is after all quantum field and not bread and butter. Plus afaik electrons have never been seen physically. They may not even be round and they may well not even behave like a spinning tennis ball lol

Spinning of electron may well be totally different to off-spin of cricket ball

Edit: In normal day to day activity we can't even control, see or feel electrons, let alone use electrons to store crysis 10 data

namnoc said:
You should be upset they haven't come up with immortality. If they had you would have plenty of time to wait for everything else.

Hehehe too damn right

panyan · 4 Jul 2011 at 23:25

Jokester said:
I would imagine just making existing silicon technology better has been so much cheaper than developing spintronics, but as you say we're now hitting a brick wall with traditional silicon semiconductors and the cost of developing new process nodes is growing massively with each reduction in manufacturing size so it makes development in alternatives more economically attractive and viable.

+1 they wont put money into alternatives until the current system becomes unviable

Judgeneo · 4 Jul 2011 at 23:28

I'll be very suprised if they move to spintronics or similar before graphene.

I reckon optical computing is the way forward myself, but haven't seen any research towards it

picktooth · 5 Jul 2011 at 06:53

what panyan said.
why devlop/introduce a new system when you can still squeeze money out of the current one.

Jaap74 · 5 Jul 2011 at 07:06

Orangey said:
As I get older I'm becoming increasingly bitter about the fact that I'm not seeing the future I was promised in science-fiction. We should have computers based off of crystals like in Stargate by now. And there aren't any flying cars either.

Are you serious ? If you're getting "bitter" I suggest you need to re-evaluate your life.......

Phenomenologica · 5 Jul 2011 at 07:43

Jaap74 said:
Are you serious ?

No, he isn't.

Yeah I was also thinking along the same line. This is after all quantum field and not bread and butter. Plus afaik electrons have never been seen physically. They may not even be round and they may well not even behave like a spinning tennis ball lol
Spinning of electron may well be totally different to off-spin of cricket ball

As I said spin itself isn't a physical spin, it's just a way of referring to it.

And in quantum theory electrons become a bit vague, they actually become an electron density field which is an area within which the electron could possibly be (Since you can't know both their momentum and position at the same time). Which I'm assuming is one of the reasons they're talking about using large numbers of electrons to do this not just one. Though I know nothing about how you measure the spin of an electron so that could be wrong. :confused:

paradigm · 5 Jul 2011 at 07:58

So "spinning" a new CPU will literally become the case

joeyjojo · 5 Jul 2011 at 10:23

WingZero30 said:
Plus afaik electrons have never been seen physically. They may not even be round and they may well not even behave like a spinning tennis ball lol
Spinning of electron may well be totally different to off-spin of cricket ball

Edit: In normal day to day activity we can't even control, see or feel electrons, let alone use electrons to store crysis 10 data

Imperial (my alma mater) recently published their results on the smoothness of electrons and found they were "surprisingly round".

The experiment, which spanned more than a decade, suggests that the electron differs from being perfectly round by less than 0.000000000000000000000000001 cm. This means that if the electron was magnified to the size of the solar system, it would still appear spherical to within the width of a human hair.

http://www3.imperial.ac.uk/newsandeventspggrp/imperialcollege/newssummary/news_26-5-2011-8-58-6

I'm not sure how that reconciles with good old wave particle duality though, as other groups are working on attosecond imaging techniques (0.000000000000000001 seconds) to "photograph" electrons in real time, and of course they're all wavy:

http://www.youtube.com/watch?v=MNrvmUvzGeM

jak731 · 6 Jul 2011 at 21:07

^ I thought that it wasn't known if electrons have any size at all (possibly point objects) and the best that can be done is give an upper bound on the size.

Orangey said:
http://www.physorg.com/news/2011-07-reveals-vital-insight-spintronics.html

Spintronics research began a long time ago. We knew CPU design would hit a wall so why have we left it so late to come up with a way forward?

iirc Spintronics is for data storage not computation. Processing power per dollar is still doubling every 18months-2 years and hard drive space has been increasing even faster than that, which is pretty incredible really.

And there aren't any flying cars either.

Helicopters?