IBM invents new liquid transistor technology

2timer

Golden Member
Apr 20, 2012
1,803
1
0
http://news.cnet.com/8301-11386_3-5...f-liquid-ibm-develops-a-new-microchip-switch/

Interesting stuff. Could someone more knowledgeable than me give some context to this quote, taken from the article:

"It's just research that's far from any actual working practical device, but Stuart Parkin, an IBM fellow at IBM Research, is willing to hold out the potential that it could help work around the difficult years ahead for Moore's Law, which faces challenges in shrinking computer circuits smaller and in getting them to run faster."

How will liquid transistors help shrink the "die size"?
 

khon

Golden Member
Jun 8, 2010
1,319
124
106
IBM's approach uses a tiny droplet of an ionic liquid electrolyte instead -- about a cubic millimeter.

A cupic milimeter might be tiny in a normal context, but in the context of transistors, it's absolutely enormous.

Current silicon based techonology could fit tens of millions of transistors in that same area.
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
The limits of silicon chips are quickly being reached. To continue the progress of Moores Law new materials will be required for building chips in the future.

Every fab on the planet is working on what comes after silicon. As of now we don't know what that is, although germanium is getting a lot of investment.
 

khon

Golden Member
Jun 8, 2010
1,319
124
106
I saw this article today which looks interesting, although it doesn't have enough details for me to say if it's realistic or not:

[url=http://www.frontline.in/stories/20130405300611000.htm]Frontline[/url] said:
A NEW breakthrough could push the limits of the miniaturisation of electronic components further than previously thought possible. A team at the Laboratoire d’Analyse et d’Architecture des Systemes (LAA), Toulouse, and Institut d’Electronique, de Microelectronique et de Nanotechnologie (IEMN) has built a nanometric transistor that displays exceptional properties for a device of its size.
To achieve this result, the researchers developed a novel three-dimensional architecture consisting of a vertical nanowire array whose conductivity is controlled by a gate measuring only 14 nm in length compared with 28 nm for the transistors in today’s chips. Published in Nanoscale, these findings offer alternatives to the planar structures used in microprocessors and memory units. The use of 3D transistors could significantly increase the power of microelectronic devices.
The “building blocks” of microelectronics, transistors consist of a semiconductor component, called channel, linking two terminals. The flow of current between these terminals is controlled by a third terminal, called gate. Acting like a switch, the gate determines whether the transistor is on or off.
It is now generally agreed that today’s transistors, with their planar architecture, are nearing the limits of miniaturisation: there is a minimum size under which the gate control over the channel becomes less and less effective. In particular, leakage currents begin to interfere with the logical operations performed by the transistor array.
The team has now built the first truly three-dimensional nanometric transistor. The device consists of a tight vertical nanowire array of about 200 nm in length linking two conductive surfaces. A chromium gate completely surrounds each nanowire and controls the flow of current, resulting in optimum transistor control for a system of this size.
This architecture could lead to the development of microprocessors in which the transistors are stacked together. The number of transistors in a given space could thus be increased considerably, along with the performance capacity of microprocessors and memory units. Another significant advantage of these components is that they are relatively simple to manufacture and do not require high-resolution lithography. In addition, these 3D transistors could be easily integrated into the conventional microelectronic devices used by the industry today.
The researchers believe that the size of the gate could be made smaller than 10 nm while still providing satisfactory control over the transistor.

Sounds like an interesting idea at least.
 

khon

Golden Member
Jun 8, 2010
1,319
124
106
When was the last time anything IBM invented went into mass production?

IBM has had the most patents every year for the past 20 years or so, literally thousands anually.

I'd say it's a pretty safe bet that all those patents are just sitting on a shelf collecting dust.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
IBM has had the most patents every year for the past 20 years or so, literally thousands anually.

I'd say it's a pretty safe bet that all those patents are just sitting on a shelf collecting dust.

Collecting dust for what? Never to be used?

When you look on IBMs financials. Its a stagnant company. IBM is good at press releases. Not so much in actual products.
 

Olikan

Platinum Member
Sep 23, 2011
2,023
275
126
Collecting dust for what? Never to be used?

When you look on IBMs financials. Its a stagnant company. IBM is good at press releases. Not so much in actual products.

so...because of this, we are not allowed to talk about this cool tecnology?
 

Idontcare

Elite Member
Oct 10, 1999
21,118
59
91
Collecting dust for what? Never to be used?

When you look on IBMs financials. Its a stagnant company. IBM is good at press releases. Not so much in actual products.

IBM is great at doing science, pushing the boundaries on the frontiers of physics. The topic in the OP is an example of that, there prolific patent activity is another.

Where IBM falls down is in the "reduction to practice" area in terms of moving one-off scientific curiosities out of the lab and into the the mass production environment of the fab.

And that isn't a new development for IBM, they have literally been like that for decades, which is why they are in the state they are in.

Liquid xtor technology is exciting because it opens up a new area of device components that can be great at doing analog stuff, including multi-regime digital (your MLC/TLC type devices), with aplomb because of the kinetics involved.

In solid-state devices the kinetics involved tend to all be deleterious (destructive) leading to degraded device parametrics over the lifetime of the device. Liquid-state devices have kinetics that can be engineered to be self-annealing, completely reversible.

That can make for efficient device operation as well as enable devices that simply would be unrealistically destructive for a true solid-state CMOS device.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
Thats how I see it too. A lot of nice cool things. But they never tend to become anything besides a press release.

I wonder if its a management fail by IBM for not materialize what their research departments work on. Or simply some kind of PR statement saying: Hey, we are here too. Please dont forget us.

Because IBM today is a ghost of its former old self. And its a slowly dying dinosaur.
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
I don't see it as a problem.

Actually I think the problem is that we no longer fund fundamental research for the sake of gaining knowledge. Now everything is about productization.
 

Homeles

Platinum Member
Dec 9, 2011
2,580
0
0
The limits of silicon chips are quickly being reached. To continue the progress of Moores Law new materials will be required for building chips in the future.

Every fab on the planet is working on what comes after silicon. As of now we don't know what that is, although germanium is getting a lot of investment.
Silicon will last us at least until the end of the decade. It's not so much of a physics problem as it is a cost problem. Quadruple patterning isn't cheap.
 

2timer

Golden Member
Apr 20, 2012
1,803
1
0
Silicon will last us at least until the end of the decade. It's not so much of a physics problem as it is a cost problem. Quadruple patterning isn't cheap.

Yes.

So thanks for the context guys, that's what I was looking for. Most of you say IBM isn't in the position to affect the transistor industry. So in all likelihood, we won't see liquid transistors for the foreseeable future.

Any thoughts on what happens below 14 nm? Will Skyline address this? I'm hearing that 3D gates are the next step.
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
Silicon will last us at least until the end of the decade. It's not so much of a physics problem as it is a cost problem. Quadruple patterning isn't cheap.

It is very much a physics problem. We are at the point where transistor design is now an exercise in quantum mechanics.
 

Idontcare

Elite Member
Oct 10, 1999
21,118
59
91
It is very much a physics problem. We are at the point where transistor design is now an exercise in quantum mechanics.

I think he means that it is not a problem wherein the solutions are limited by physics. All our current problems are limited in solution only by economics for the foreseeable decade or two.

Yes the money goes towards enabling solutions that are based on physics, but not verboten per physics.
 

Hulk

Diamond Member
Oct 9, 1999
4,375
2,255
136
Perhaps the onus of advancement will have to shift from the current paradigm of hardware leading software to software leading the hardware. Or more appropriately working with the hardware available. Instead of coding for tomorrow with the thought that by the time it really hits the streets the hardware will be there to support it, it might be time to rethink software development.

Perhaps some hand coded assembly and a real effort to code for the special instructions and strengths of current processor designs. The increases in computer system performance from processors to memory to storage to graphics may have allowed programmers to get a little bit lazy.

Back in high school I used to fool around with game programming on my, ahem, Atari 800. It was all Assembly and believe me you can do a lot with 4MHz in Assembly. In the 160x192 resolution mode, 2 bits per pixel (4 color) there was enough time during the vertical blank interrupt (the time electron beam turned off at the end of one scan line and before it started to draw the next) we could do things like change color registers. So in horizontal sections down the screen you could have as many colors as you wanted. Just not more than 4 in a single scan line.

Another example. In the early '90's I used a multitrack audio program called SAW which was hand coded by Bob Lentini. It was all of 4MB and ran from the exe. It was about 10 times faster than anything available at the time. Or probably to this day. Imagine 8 tracks of audio with eq, compression etc.. on a Pentium 60. Real time.

If you need to find the area under a function you can be lazy and use compute with the trapezoidal rule. Or you can be clever and integrate the function. Too much trapezoidal rule these days I fear.

Time to think smarter and not just faster.
 

Extelleron

Diamond Member
Dec 26, 2005
3,127
0
71
Thats how I see it too. A lot of nice cool things. But they never tend to become anything besides a press release.

I wonder if its a management fail by IBM for not materialize what their research departments work on. Or simply some kind of PR statement saying: Hey, we are here too. Please dont forget us.

Because IBM today is a ghost of its former old self. And its a slowly dying dinosaur.

Is everyone else in this thread looking at the same IBM I am looking at? IBM's net income in 2012 was $16.6 billion. Net income has increased every year since 2003. Efficiency way up, op margin has gone from 37.7% in 2003 to 49.1% in 2012.

Last year Intel made $11 billion, down from the previous year. And only ~10% higher than net income in 2000, 12 years ago.

IBM's equity is worth $236B, Intel's is worth $105B. On a price/earnings basis, investors are willing to pay roughly twice for IBM compared to what they would pay for Intel - signaling that they feel a lot more comfortable about IBM's future prospects than they do Intel's.

Just because IBM doesn't have products which seem exciting to you doesn't mean they aren't building products businesses are using and willing to pay a lot of money for.
 
Last edited:

Wall Street

Senior member
Mar 28, 2012
691
44
91
Is everyone else in this thread looking at the same IBM I am looking at? IBM's net income in 2012 was $16.6 billion. Net income has increased every year since 2003. Efficiency way up, op margin has gone from 37.7% in 2003 to 49.1% in 2012.

You beat me to it. Tons of people running their mouth here who don't know what a 10-K is.

As for this news, the issue here is always mass production and cost. While there have been at least 100 "breakthroughs" that I have seen in research papers that beat the current industry standard of silicon lithography, a vast majority are fail when it comes to reliably making billion+ transistor chips that need to cost <$100. Input materials costs, low yields and overly complicated/long processes can all lead to lack of economic viability.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
Is everyone else in this thread looking at the same IBM I am looking at? IBM's net income in 2012 was $16.6 billion. Net income has increased every year since 2003. Efficiency way up, op margin has gone from 37.7% in 2003 to 49.1% in 2012.

Last year Intel made $11 billion, down from the previous year. And only ~10% higher than net income in 2000, 12 years ago.

IBM's equity is worth $236B, Intel's is worth $105B. On a price/earnings basis, investors are willing to pay roughly twice for IBM compared to what they would pay for Intel - signaling that they feel a lot more comfortable about IBM's future prospects than they do Intel's.

Just because IBM doesn't have products which seem exciting to you doesn't mean they aren't building products businesses are using and willing to pay a lot of money for.

http://seekingalpha.com/article/1085441-is-ibm-in-trouble
http://dividendmonk.com/international-business-machines-ibm-solid-value/
http://articles.businessinsider.com...-contract-project-management-business-insider

Income just changed because they went from hardware to software and could shrink the organization. One day they cant keep firing people to boost income.
 
Last edited:

Cogman

Lifer
Sep 19, 2000
10,278
126
106
http://seekingalpha.com/article/1085441-is-ibm-in-trouble
http://dividendmonk.com/international-business-machines-ibm-solid-value/
http://articles.businessinsider.com...-contract-project-management-business-insider

Income just changed because they went from hardware to software and could shrink the organization. One day they cant keep firing people to boost income.

IBM dumped consumer level hardware, however their Enterprise level stuff is still alive, well, and pulling in a decent chunk of money for them.

Yes, they are somewhat stagnant, however they aren't really in pain either. They are still a giant in the room with income that dwarfs their nearest competitor. Their current market cap is at 236.37B, Intel, by comparison, is at 105.50B.

There are very few companies out there that have a higher market cap.
 

Extelleron

Diamond Member
Dec 26, 2005
3,127
0
71
http://seekingalpha.com/article/1085441-is-ibm-in-trouble
http://dividendmonk.com/international-business-machines-ibm-solid-value/
http://articles.businessinsider.com...-contract-project-management-business-insider

Income just changed because they went from hardware to software and could shrink the organization. One day they cant keep firing people to boost income.

So you're arguing IBM has managed to artificially boost their income every single year for the past decade solely by cutting employees? And further, that the fact that income has been artificially boosted and will eventually come crashing down has gone unnoticed by millions of investors, including Warren Buffett (who has aggressively bought IBM for Berkshire and now owns about 6% of the company), but is so obvious that you can prove it with one line and a link to a post on Seeking Alpha?
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
So you're arguing IBM has managed to artificially boost their income every single year for the past decade solely by cutting employees? And further, that the fact that income has been artificially boosted and will eventually come crashing down has gone unnoticed by millions of investors, including Warren Buffett (who has aggressively bought IBM for Berkshire and now owns about 6% of the company), but is so obvious that you can prove it with one line and a link to a post on Seeking Alpha?

So because Warren buys stocks it all good? Thats what your argument is.

Raised profit is always good for investors. IBM just moved to higher margin fields with software. But its layoffs and outsourcing. IBM havent had any real growth, its simply stagnant. Corrected for inflation its actually shrinking.
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
IBM havent had any real growth, its simply stagnant. Corrected for inflation its actually shrinking.

Outside of 2012 IBM has growing both the top an bottom lines at about 7% a year since 2008.

What's the inflation rate in Denmark? It's been about 2% in the US during the same time frame.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |