Download the Tom's Hardware App from the App Store
The reference for current tech news
Yes No
Ads

Lasers Enable Finer Chip Structures, Advance of Moore's Law

by - source: MIT

A research team at the Massachusetts Institute of Technology (MIT) said it has found a new way to shrink circuit structures in semiconductors.

Current chip manufacturing relies on photolithography techniques, which can only produce chip features that are larger than the wavelength of the light applied.

A process that is described in the paper "Breaking the Far-Field Diffraction Limit in Optical Nanopatterning via Repeated Photochemical and Electrochemical Transitions in Photochromic Molecules", published in Physical Review Letters, can create complex structures chip structures 1/8 the size of the wavelength of the light used. According to the researchers, an effect called stimulated emission depletion imaging (STED) enabled them to go beyond the current limitations of photolithography. In STED, scientists leverage the fluorescent characteristics of materials to emit light when targeted by a laser beam. By controlling the laser's power, the researchers can affect the strength of light emitted and, if the power falls enough, cause a "dark patch" that is smaller than the wavelength of the laser light itself. These dark patches can be used as masks, which can be applied to a surface.

The MIT researchers believe that their invention could be used to create semiconductors with much finer structures than possible today. There could also be an opportunity to apply this technology in photonic devices, MIT said.

Share:
68
Comments
X
Submit

Comments
Add your comment
lol92 12/26/2011 2:07 PM
Hide
-16+

really? that good. i hope we can see a lower than 10nm architecture

theuniquegamer 12/26/2011 2:12 PM
Hide
-4+

That means we can get 32 or more cores in the mobile chips in next few years.

amk-aka-phantom 12/26/2011 2:16 PM
Hide
-20+

theuniquegamer :
That means we can get 32 or more cores in the mobile chips in next few years.



... and over 9000 in the desktop ones :D

Remember, whatever "mobile" has to offer, desktop will always pwn it.

guru_urug 12/26/2011 2:20 PM
Hide
-6+

Combine this with the molybdenite tech in the previous article and we have pocket supercomputers. Seriously the future is really exciting!!! Cant wait

salgado18 12/26/2011 2:24 PM
Hide
-20+

Now THAT deserves a patent, not some buttons on scren.

amk-aka-phantom 12/26/2011 2:34 PM
Show
salgado18 12/26/2011 2:41 PM
Hide
-20+

amk-aka-phantom wrote :

No way. You're again patenting an IDEA, FFS! That will just slow the implementation down! Where would the humanity be if everyone would be patenting everything from the invention of the wheel, huh?



That's real technology, a new process of production of really small circuits. It takes years of research, a lot of knowledge and study, and a hell lot of money to even make it right, so it definitely deserves a patent to these guys.

If that's not a patent, then what is?

Anonymous 12/26/2011 2:44 PM
Hide
-6+

I like this.

mildgamer001 12/26/2011 2:47 PM
Hide
-10+

amk-aka-Phantom :
... and over 9000 in the desktop ones Remember, whatever "mobile" has to offer, desktop will always pwn it.


ITS OVER 9000!!!!

De5_roy 12/26/2011 2:54 PM
Hide
-7+

over 9000!!!!!

bgaimur 12/26/2011 3:06 PM
Hide
-11+

brb, going in my time machine 50 years from now. See you guys later :D

wiyosaya 12/26/2011 3:21 PM
Hide
-9+

lol92 wrote :

really? that good. i hope we can see a lower than 10nm architecture




Here is another report on the article that states things in what is, IMHO, a more accurate manner. It implies that the technology is capable of producing features that are 1/8 the size of what is currently used in the industry with a technology that is completely compatible with existing semiconductor fabrication technology.

IMHO, it is potentially more significant than the Tom's article implies - if, of course, it is actually employed in the industry.

theuniquegamer wrote :

That means we can get 32 or more cores in the mobile chips in next few years.



So here's the cool thing about this, IMHO. I think the "core count" will scale as a function of area (i.e., a square function). So, for 1/8 current dimensions, you have 64x the core count. Which could mean 512 cores in the very near future. :wahoo:

amk-aka-phantom 12/26/2011 3:26 PM
Show
uproller 12/26/2011 3:31 PM
Hide
-15+

amk-aka-Phantom :
I really don't care. Patents = slow-downs in technical progress. I'm a selfish bastard and don't care for anything but my benefit If their research was not for the benefit of the mankind but for money, they can go #@$! themselves, period.



Oh the irony.

stingstang 12/26/2011 5:02 PM
Hide
-9+

wiyosaya :
Here is another report on the article that states things in what is, IMHO, a more accurate manner. It implies that the technology is capable of producing features that are 1/8 the size of what is currently used in the industry with a technology that is completely compatible with existing semiconductor fabrication technology.IMHO, it is potentially more significant than the Tom's article implies - if, of course, it is actually employed in the industry.So here's the cool thing about this, IMHO. I think the "core count" will scale as a function of area (i.e., a square function). So, for 1/8 current dimensions, you have 64x the core count. Which could mean 512 cores in the very near future.

say Imho one more time..

freggo 12/26/2011 5:05 PM
Hide
-11+

amk-aka-Phantom :
I really don't care. Patents = slow-downs in technical progress. I'm a selfish bastard and don't care for anything but my benefit If their research was not for the benefit of the mankind but for money, they can go #@$! themselves, period.



Hmm, may I ask then what have YOU done for mankind ? :-)
Or do you just expect others to do for you ?

You can not expect a business to do research etc for free. After all they have to somehow pay their researchers, the research assistants, the staff and the janitor. Hell, YOU may be one of the people they have to pay.
there is nothing wrong with making a buck and protecting your investment of time and brainpower if it is within socially responsible limits.

amk-aka-phantom 12/26/2011 5:08 PM
Show
amk-aka-phantom 12/26/2011 5:09 PM
Show
amk-aka-phantom 12/26/2011 5:09 PM
Show
Inanition02 12/26/2011 5:21 PM
Hide
-15+

amk-aka-Phantom :
Tesla did it. Did more than anyone else at his time.



Tesla often worked for profit - and created companies to sell his inventions. Also, we asked if you did it, not someone else. Bill Gates gives away billions - does that mean you're charitable? Or that everyone should give away everything?

Also, if you want to be Tesla, go ahead - "Because of his eccentric personality and his seemingly unbelievable and sometimes bizarre claims about possible scientific and technological developments, Tesla was ultimately ostracized and regarded as a mad scientist by many late in his life.[7] He died without much money to his name.[8]"

Inanition02 12/26/2011 5:23 PM
Hide
-14+

amk-aka-Phantom :
Ever heard of open source?



How many people make a living from open source companies that don't protect/sell anything at all?? Oh...right...sure some companies give away some things - but no one gives away everything, at least not without a way to monetize it in some way (support, ads, licensing, long term stable versions, etc).

Anyhow, do you create large innovative open source software for free?

amk-aka-phantom 12/26/2011 5:45 PM
Show
memadmax 12/26/2011 5:55 PM
Hide
-3+

With all these advances in processor technology, skynet should be online shorty...

/rubshands =D

cookoy 12/26/2011 6:39 PM
Hide
-5+

Granting a patent to this new process technology would take a lot longer as the clerk would need a week just to understand each word in the title. But just about anyone can approve a rectangular design.

freggo 12/26/2011 6:55 PM
Hide
-7+

amk-aka-Phantom :
Ever heard of open source?



What has hardware development to do with open source ?


freggo 12/26/2011 7:03 PM
Hide
-4+

amk-aka-Phantom :
Tesla did it. Did more than anyone else at his time.



Yes, that was a long time ago when research was a 'one man job' and you could to it on a research grant to gain personal fame ( or one woman of course as in Madam Curie ).

In the 70s I wrote video games. Took typically 2-3 month for a game written in Assembler. You could do that 'open source' or just for the fun of it.
Todays top games are made by a team of programmers at a cost well above $100.000.000.

Same with hardware, the days of Tesla and Einstein are over. Research is a team effort using hugely expensive hardware etc. Unless it is government sponsored you have to make a buck at the end to get investors.
Of course, if your research is weapons related you may get funding from various non-disclosed agencies with typically 3 letter names ;-)

amk-aka-phantom 12/26/2011 7:08 PM
Show
JohnnyLucky 12/26/2011 7:56 PM
Hide
-2+

How would applications be affected if this technology became a reality?

amk-aka-phantom 12/26/2011 8:19 PM
Show
cmcghee358 12/26/2011 8:27 PM
Hide
-6+

amk-aka-Phantom :
No. I deploy open source software on the large scale for free. Almost as much pain in the ass, if not more, trust me I know who he was, no need to copy Wikipedia. It's fine with me; the point of life is doing something that you enjoy, not "dying with much money to your name". That is a perfect consumerist's dream - get a family, kids, pay bills, grow old, die. Not bad things, but certainly not the main goal.Before you swindle me into further off-topic, again, I do not care who and for how long has been developing what. Patents hold development back, and that is an undeniable fact.Selling is fine. Patent trolling isn't, and every patent ends up like that nowadays. It's fine to patent a particular model or architecture - needs to be defined very precise, though, not "use CPU to process input" or whatever Apple comes up with in their patents - not a generalized idea.



If you think that patents hold us back, as a society, you are quite stupid.

The simple fact is, the patent is what drives and motivates technological improvement. Because the inventors know their creation will 1. be secure and 2. be profitable; it drives people to create new innovations.

Where is the motivation if you knew you wouldn't make any money off your life's work.

Herp derp derp. Society is hard to understand.

amk-aka-phantom 12/26/2011 8:49 PM
Show

Ads

Best offers

Newsletters


OK
Ads