mailing list For webmasters
Welcome Guest
Now, a seven-atom transistor Options
kisholoy mukherjee
Posted: Tuesday, May 25, 2010 2:58:28 PM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
SYDNEY: Australian scientists unveiled the world's smallest electronic switch measuring just a few atoms, which will shrink microchips and revolutionize computing speeds.

http://timesofindia.indiatimes.com/home/science/Seven-atom-transistor-spells-quantum-leap-for-computing/articleshow/5970558.cms
Luftmarque
Posted: Tuesday, May 25, 2010 3:19:02 PM

Rank: Advanced Member

Joined: 3/17/2009
Posts: 3,119
Neurons: 39,962
Location: Pau, Aquitaine, France
Very nice! Smaller = faster = lower power consumption = cooler. Besides which, as a quantum computer, it may work on NP-complete problems? So… the positronic brain is a mere decade away then. Hopefully more like Star Trek's Data than Battlestar Galactics's Cylons.
nooblet
Posted: Tuesday, May 25, 2010 3:33:24 PM
Rank: Advanced Member

Joined: 10/15/2009
Posts: 1,570
Neurons: 4,778
Location: United States
Amen to that Mark.

I am astonished that a transistor was reduced down to 7 atoms... That is insane. I can't read the article at the moment, but the transistor is specifically for quantum computing? It holds 3 states? If that is the case, I believe there should also be one that is developed for 2 states, since quantum computing has some serious deficiencies where normal computers can handle the solutions elegantly. Each one has their strengths and weaknesses, and if transistors are getting down to 7 atoms, it should be possible to run a quantum and a binary setup in parallel.
kisholoy mukherjee
Posted: Tuesday, May 25, 2010 4:13:00 PM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
I would be interested to know if it follows the same p-n-p or n-p-n configuration. Not much has been said in that article about the procedure for fabricating this transistor.
But I can understand it having only seven atoms, since majority and minority carries in holes and electrons are all that we need in a transistor.

nooblet
Posted: Tuesday, May 25, 2010 4:52:00 PM
Rank: Advanced Member

Joined: 10/15/2009
Posts: 1,570
Neurons: 4,778
Location: United States
But it's only 7 atoms... I dunno, the fact that anything man-made can be that small is pretty impressive. I'm honestly not all that familiar with how a transistor works, except that it is often used as a switch. How they are capable of acting as switches is something I don't fully understand, although I know it has to do with having part of the transistor being saturated with electrons.
uuaschbaer
Posted: Tuesday, May 25, 2010 4:53:30 PM

Rank: Advanced Member

Joined: 10/18/2009
Posts: 1,928
Neurons: 6,180
And this on Towel Day! Which has a vague, far-fetched kind of relevance.

"Today you can carry a computer around in your hand and many of its components are more than 1000 times smaller than the width of a human hair.

I didn't know this (the second part of this obviously). Does anybody know which components are so small? Noteworthy by the way is that Moore's law was about transistors rather than memory but oddly enough it seems to work for anything that people mistake it to have been about.
nooblet
Posted: Tuesday, May 25, 2010 5:01:24 PM
Rank: Advanced Member

Joined: 10/15/2009
Posts: 1,570
Neurons: 4,778
Location: United States
They're talking about transistors, aka the microswitches inside the CPU.
kisholoy mukherjee
Posted: Tuesday, May 25, 2010 5:10:52 PM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
nooblet wrote:
But it's only 7 atoms... I dunno, the fact that anything man-made can be that small is pretty impressive. I'm honestly not all that familiar with how a transistor works, except that it is often used as a switch. How they are capable of acting as switches is something I don't fully understand, although I know it has to do with having part of the transistor being saturated with electrons.


It is essentially similar to the diode, except that it has amplifying properties whereas the diode mainly acts as a switch or a converter (rectifier..AC to DC).
The transistor has three parts; the emitter, which as the name suggests, emits majority carriers (electrons or holes), base, which connects the emitter to the collector, and the collector. (Emitter-collector voltage = 0)
It comes as n-p-n or p-n-p. P is where there are more holes (or less electrons) and n is where there are more electrons.

All the layers have the semiconductor doping, and the doping is of varying levels.
The electrons or holes can travel through one layer to the next (say emitter to the base) because of concentration gradient of carriers. The part where 'saturation' takes place would be the depletion layer, which then inhibits the passage of electrons and holes because of recombination.
You can forward or reverse bias the transistor (forward: positive end of battery to p-junction and reverse : just the opposite) Forward bias decreases the depletion layer while reverse bias increases the depletion region.

The applied voltage across the transistor terminals has to be greater than a certain breakdown voltage, beyond which the carriers will start flowing across the junctions and hence current will flow. So, by regulating the voltage to the terminals, we can use the transistor as a switch.
kisholoy mukherjee
Posted: Tuesday, May 25, 2010 5:17:59 PM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
nooblet wrote:
They're talking about transistors, aka the microswitches inside the CPU.


Indeed, I just found out this piece of info:

The hair strands are "17 to 181 µm" in average. While the smallest transistors inside the Intel Pentium 4 (which has millions of transistors) are of 40- 100 nm (nano meters).
So that ratio indeed roughly comes to around 1000.
uuaschbaer
Posted: Tuesday, May 25, 2010 5:25:01 PM

Rank: Advanced Member

Joined: 10/18/2009
Posts: 1,928
Neurons: 6,180
nooblet wrote:
They're talking about transistors, aka the microswitches inside the CPU.


They do but in the last paragraph they mention memory. This makes you first think that Moore's law isn't relevant to the report of the transistor but it is. Well RAM memory has something to do with transistors perhaps, I don't know.
nooblet
Posted: Tuesday, May 25, 2010 5:33:27 PM
Rank: Advanced Member

Joined: 10/15/2009
Posts: 1,570
Neurons: 4,778
Location: United States
Yes, transistors are in RAM (and any integrated circuit), but I was actually referring to your question about which piece of a computer is 1/1000 the size of the width of a human hair. Transistors are that small.
kisholoy mukherjee
Posted: Tuesday, May 25, 2010 5:42:26 PM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
nooblet wrote:
Yes, transistors are in RAM (and any integrated circuit), but I was actually referring to your question about which piece of a computer is 1/1000 the size of the width of a human hair. Transistors are that small.


Yes, there can be diodes and transistors in RAM. Actually, the smallest unit of memory is a flip flop or a latch. There are many types, the most common one being a D flip flop. It simply 'holds' or 'passes' a single bit of information, using certain signals like 'Enable'.
When you have a 4 bit memory device or a register, it means that there are 4 1-bit flip flops in it.
And the Flip flop is obviously made of analog components like resistors, transistors etc.
uuaschbaer
Posted: Tuesday, May 25, 2010 5:55:39 PM

Rank: Advanced Member

Joined: 10/18/2009
Posts: 1,928
Neurons: 6,180
nooblet wrote:
Yes, transistors are in RAM (and any integrated circuit), but I was actually referring to your question about which piece of a computer is 1/1000 the size of the width of a human hair. Transistors are that small.


O deary me, how rude of me. I'm sorry. Boy, those transistors are small. I shall be looking out for a low-budget show on Discovery to see how that is made.
kisholoy mukherjee
Posted: Tuesday, May 25, 2010 6:00:43 PM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
uuaschbaer wrote:
nooblet wrote:
Yes, transistors are in RAM (and any integrated circuit), but I was actually referring to your question about which piece of a computer is 1/1000 the size of the width of a human hair. Transistors are that small.


O deary me, how rude of me. I'm sorry. Boy, those transistors are small. I shall be looking out for a low-budget show on Discovery to see how that is made.


Check this out.
http://101science.com/transistor.htm#FABRICATION
http://www.youtube.com/watch?v=yMyR_3odegM
uuaschbaer
Posted: Tuesday, May 25, 2010 6:09:03 PM

Rank: Advanced Member

Joined: 10/18/2009
Posts: 1,928
Neurons: 6,180
kisholoy mukherjee wrote:
uuaschbaer wrote:
nooblet wrote:
Yes, transistors are in RAM (and any integrated circuit), but I was actually referring to your question about which piece of a computer is 1/1000 the size of the width of a human hair. Transistors are that small.


O deary me, how rude of me. I'm sorry. Boy, those transistors are small. I shall be looking out for a low-budget show on Discovery to see how that is made.


Check this out.
http://101science.com/transistor.htm#FABRICATION
http://www.youtube.com/watch?v=yMyR_3odegM


Very nice. I always wonder how they're going to find back what they have produced under a microscope and make sure it doesn't fly out of the window on a gust of wind. I'll save the technical documentation for tomorrow as I can't afford to stay up right now. Thanks massively, though.
kaleem
Posted: Tuesday, May 25, 2010 6:50:43 PM
Rank: Advanced Member

Joined: 12/27/2009
Posts: 3,252
Neurons: 9,948
Very nice - good things come in small packaging!




kisholoy mukherjee
Posted: Tuesday, May 25, 2010 8:54:23 PM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
uuaschbaer wrote:
I'll save the technical documentation for tomorrow as I can't afford to stay up right now. Thanks massively, though.


These are better.
http://www.circuitstoday.com/ic-fabrication-techniques
http://www.circuitstoday.com/silicon-substrate-preparation
worldsclyde
Posted: Tuesday, May 25, 2010 9:05:52 PM
Rank: Advanced Member

Joined: 4/24/2010
Posts: 785
Neurons: 2,390
Location: Spokane, WA USA
I'm curious if the transistor was stable or if it only functioned for a short time.
kisholoy mukherjee
Posted: Tuesday, May 25, 2010 9:14:48 PM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
worldsclyde wrote:
I'm curious if the transistor was stable or if it only functioned for a short time.


I think it is stable or at least it can be made (more) stable during industry level manufacturing which will not have the constraints faced in research level.
Plus, it has been mentioned in the article that it can be useful for quantum computing within five years and that it is 'compatible' with existing technology, so I think we can assume that the technology is stable.
uncoverer
Posted: Wednesday, May 26, 2010 12:36:42 AM
Rank: Advanced Member

Joined: 1/13/2010
Posts: 129
Location: USA
nooblet:it may work on NP-complete problems?


That is an interesting idea nooblet which nobody touched upon in the subsequent discussions.
We can only wonder though whether it will work on np complete problem.
kisholoy mukherjee
Posted: Wednesday, May 26, 2010 4:08:14 AM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
Luftmarque wrote:
Besides which, as a quantum computer, it may work on NP-complete problems?


The article says that problems which would have taken more time than the age of the universe itself with ordinary computers can be solved by such a computer. Does that include Np complete problems also? I don't know of NP-complete problems. Will you please explain what they are?
michiko
Posted: Wednesday, May 26, 2010 7:16:23 AM
Rank: Advanced Member

Joined: 4/25/2010
Posts: 44
Neurons: 131
kisholoy mukherjee wrote:
SYDNEY: Australian scientists unveiled the world's smallest electronic switch measuring just a few atoms, which will shrink microchips and revolutionize computing speeds.

http://timesofindia.indiatimes.com/home/science/Seven-atom-transistor-spells-quantum-leap-for-computing/articleshow/5970558.cms



its freakin' amazing!...another leap for humankind. job very well done Australian ScientistsApplause
kisholoy mukherjee
Posted: Wednesday, May 26, 2010 7:28:38 AM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
michiko wrote:
kisholoy mukherjee wrote:
SYDNEY: Australian scientists unveiled the world's smallest electronic switch measuring just a few atoms, which will shrink microchips and revolutionize computing speeds.

http://timesofindia.indiatimes.com/home/science/Seven-atom-transistor-spells-quantum-leap-for-computing/articleshow/5970558.cms



its freakin' amazing!...another leap for humankind. job very well done Australian ScientistsApplause


Oh I forgot to mention...there was an Indian in the team as well. Sudhashatta Mahapatra. Probably from Orissa, India. So hoorah for him tooDancing


[image not available]
Luftmarque
Posted: Wednesday, May 26, 2010 8:51:50 AM

Rank: Advanced Member

Joined: 3/17/2009
Posts: 3,119
Neurons: 39,962
Location: Pau, Aquitaine, France
kisholoy mukherjee wrote:
Luftmarque wrote:
Besides which, as a quantum computer, it may work on NP-complete problems?


The article says that problems which would have taken more time than the age of the universe itself with ordinary computers can be solved by such a computer. Does that include Np complete problems also? I don't know of NP-complete problems. Will you please explain what they are?

There is a class of mathematical problems that are called Non-Polynomial (time) Complete. Some of them in crypology. The equations to solve them involve functions other than polynomial (like c + x + y^2 - z^3). Exponential terms appear. The class is called Complete because it has been shown that solving any of them will also solve all the others. I think that's it (I'm on my phone in the back yard & it would be annoying to Google it right now).
michiko
Posted: Wednesday, May 26, 2010 8:55:27 AM
Rank: Advanced Member

Joined: 4/25/2010
Posts: 44
Neurons: 131
haha! that's enlightening KM, hoorah for him as well! Applause
kisholoy mukherjee
Posted: Wednesday, May 26, 2010 9:29:09 AM
Rank: Advanced Member

Joined: 10/31/2009
Posts: 3,729
Neurons: 7,777
Location: here and there
michiko wrote:
haha! that's enlightening KM, hoorah for him as well! Applause


Yes indeed. :)
and thanks to Mark for explaining that part.
uuaschbaer
Posted: Saturday, May 29, 2010 10:03:56 AM

Rank: Advanced Member

Joined: 10/18/2009
Posts: 1,928
Neurons: 6,180
kisholoy mukherjee wrote:
uuaschbaer wrote:
I'll save the technical documentation for tomorrow as I can't afford to stay up right now. Thanks massively, though.


These are better.
http://www.circuitstoday.com/ic-fabrication-techniques
http://www.circuitstoday.com/silicon-substrate-preparation


They are, they're very good. Very detailed.
Users browsing this topic
Guest


Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.