Intel announces Tri-gate "3-D" transistors for upcoming Ivy Bridge based processors

[AdamK47]

Apple rumors = page hits

Lets see if I can drum up some Google search hits to this thread.

Apple buys Intel
Intel develops Apple CPU
Apple Intel CPU chip 22nm Deal
Intel Apple Merger
Apple Stock
Intel Stock

 

[AdamK47]

Whys that? its a 3D chip. and pretty much somes up its design in its name. Planar sounds better to you some how .

Sum up, yes. Some up, no... well, maybe.

Nothing sounds better to me. Perhaps your telepathic board reading powers got crossed with someone else.

On second thought, 4D transistors sounds better to me. Who needs these lame 3D transistors when you have 4D.

 

[OCGuy]

Lets see if I can drum up some Google search hits to this thread.

Apple buys Intel
Intel develops Apple CPU
Apple Intel CPU chip 22nm Deal
Intel Apple Merger
Apple Stock
Intel Stock

 

[Nemesis 1]

Sum up, yes. Some up, no... well, maybe.

Nothing sounds better to me. Perhaps your telepathic board reading powers got crossed with someone else.

On second thought, 4D transistors sounds better to me. Who needs these lame 3D transistors when you have 4D.

dang you quoted befor i seen that. and fixed it . Now what is 4D. quantuam

 

[TuxDave]

Although it's confusing because literature and the media often use FinFet to mean any non-planar device, and frequently use the terms interchangeably, the original FinFET design from UC Berkeley is a non-planar ("3D") dual-gate transistor. Intel's tri-gate design is a non-planar tri-gate transistor.

The paper that I consider to be the one that named the "finFET" is this one:
"Sub-50nm P-channel FinFET" Quejue Huang et.al, IEEE Transactions on Electron Devices, Vol. 48, No.5, May 2001.
In the abstract: "High-performance PMOSFETs with sub-50-nm gate length are reported. A self-aligned double-gate MOSFET structure (FinFET) is used to suppress the short-channel effects."

If you Google it, you can read the paper. Like I said, I find it confusing too - first because a finFET looks like a tri-gate when you look at the electron microscope photos, and second because people seem to use the term "finFET" to mean any non-planar multigate transistor. To me, when I hear the term "finFET" I think of the paper from 10 years ago and then the subsequent dual-gate papers from TSMC and this is a distinctly different design from Intel's tri-gate.

* Not a spokesperson for Intel Corp. *

It could be because I'm a Cal Alumni that I fall under the pile of people that use "finFETs" to mean any non-planar multigate transistor. It is different in fabrication and physical properties.... BUT... I'd like to say it goes along the same principles where a non-planar transistor will have better control over the channel.

 

[Soleron]

The 37% increase in performance if translates to clocks kinda takes the fun out of debating IB clocks speeds. That should put IB at about 4.30ghz at stock clocks without turbo

Anandtech's article says that 37% refers to lower voltages and it's only 18% at the high end. 4GHz then. Intel will likely use that to instead lower per-core power so they can have more cores, so I don't think we'll even see all of that increase as clocks.

 

[TuxDave]

More surface area for a single drain+source? (depends on how thick the fin is?).

I can't really speculate on the details but in the traditional sense of finfets, the thickness of the FINs are usually set and increased drive current is achieved by using multiple fins (increasing the total surface area between source/drain to increase current).

There is a lot of goodness in the technology but as far as I recall, the #1 bonus was better channel control. If you have a picture of a planar transistor, the gate controls the channel near the top and so it has really good control for the channel nearest to the surface. But as you go deeper into the substrate it's less effective.

With a non-planar channel, you have basically gates all around the channel and so you can better control leakage and get all the other goodness.

 

[Nemesis 1]

Anandtech's article says that 37% refers to lower voltages and it's only 18% at the high end. 4GHz then. Intel will likely use that to instead lower per-core power so they can have more cores, so I don't think we'll even see all of that increase as clocks.

No disrespect but I think Ananda has it right and wrong. Intel can do both and likely will.

 

[AdamK47]

dang you quoted befor i seen that. and fixed it . Now what is 4D. quantuam

Nah, it's too early for Intel to release their quantum transistors they have been working on. After looking at the picture on the BBC site, I actually think they should call it waffle transistor technology. 22nm waffle transistors.

 

[piesquared]

the news that the intel marketing mushrooms are trying to hide, is that 22nm schedule has slipped from 'delivery in the 2nd half of 2011' to volume production by the end of the year. of course those who trolled about Glofo's 32nm being pushed back will simply ignore this fact or go whine and cry to the moderators. but i never really expected anything different here.



Keep up the trolling and you won't have to worry about what's being posted in this forum.

Anandtech Admin
Red Dawn

 

[Phynaz]

Zoinks Scooby!

I don't know how Intel does it, this is the second fundamental change in transistor design they have made in a couple of years.

Somebody is smiling down on them.

 

[Nemesis 1]

the news that the intel marketing mushrooms are trying to hide, is that 22nm schedule has slipped from 'delivery in the 2nd half of 2011' to volume production by the end of the year. of course those who trolled about Glofo's 32nm being pushed back will simply ignore this fact or go whine and cry to the moderators. but i never really expected anything different here.

intel only said we well see 22nm chips in 2011. they never said what product. AS long as IB is 1stqt 2012 none care. If you have a link saying IB in 2011 lets see it. It must come from intel .

 

[Nemesis 1]

Zoinks Scooby!

I don't know how Intel does it, this is the second fundamental change in transistor design they have made in a couple of years.

Somebody is smiling down on them.

ya think?

 

[PlasmaBomb]

Anandtech's article says that 37% refers to lower voltages and it's only 18% at the high end. 4GHz then. Intel will likely use that to instead lower per-core power so they can have more cores, so I don't think we'll even see all of that increase as clocks.

Lower per core power likely means higher maximum possible turbo though.

 

[(sic)Klown12]

Zoinks Scooby!

I don't know how Intel does it, this is the second fundamental change in transistor design they have made in a couple of years.

Somebody is smiling down on them.

Being the industry leader and having an incredible amount of money to throw at problems has it's advantage. They really are an amazing company full of smart and competent people.

 

[Phynaz]

intel only said we well see 22nm chips in 2011. they never said what product. AS long as IB is 1stqt 2012 none care. If you have a link saying IB in 2011 lets see it. It must come from intel .

Please don't feed the troll, just report him.

 

[Nemesis 1]

Please don't feed the troll, just report him.

I was going to do the troll troll troll your boat gently down the stream . But figured the mods would see it . and they did. I used that report link earlier this year several times didn't like using it. SO I stopped

 

[StinkyPinky]

The next year is going to be really interesting for CPU's. There are going to be some absolute beasts being released, even in the mid-end.

Which is good because GPU's are becoming boring.

 

[Majic 7]

I have what I think is a perfect system now that I have my B3 board, haven't had a single problem with it except I had to reinstall OS to fix a sleep problem. Since then it has been the most problem free system I have ever had. Easy overclock, no gaming problems, no system niggling why did it do that problems, hardly any event viewer notifications, nothing. Now this is coming, crap!

 

[Hogan773]

The next year is going to be really interesting for CPU's. There are going to be some absolute beasts being released, even in the mid-end.

Which is good because GPU's are becoming boring.

And you'll be saying the same thing next year. There will be some absolute beasts coming in 2013.

 

[Vesku]

Looks like Intel is getting ready to up the core counts for 22nm, guess we'll see if the programmers are ready to take advantage.

Is this May information release a typical time for Intel or is the timing a bit of marketing before Llano goes retail?

 

[nyker96]

the Ivy looks pretty awesome, 1/2 power consumption too! glad I skipped the current gen stuff. 22nm intel will be sweet!

 

[theeedude]

Although it's confusing because literature and the media often use FinFet to mean any non-planar device, and frequently use the terms interchangeably, the original FinFET design from UC Berkeley is a non-planar ("3D") dual-gate transistor. Intel's tri-gate design is a non-planar tri-gate transistor.

The paper that I consider to be the one that named the "finFET" is this one:
"Sub-50nm P-channel FinFET" Quejue Huang et.al, IEEE Transactions on Electron Devices, Vol. 48, No.5, May 2001.
In the abstract: "High-performance PMOSFETs with sub-50-nm gate length are reported. A self-aligned double-gate MOSFET structure (FinFET) is used to suppress the short-channel effects."

If you Google it, you can read the paper. Like I said, I find it confusing too - first because a finFET looks like a tri-gate when you look at the electron microscope photos, and second because people seem to use the term "finFET" to mean any non-planar multigate transistor. To me, when I hear the term "finFET" I think of the paper from 10 years ago and then the subsequent dual-gate papers from TSMC and this is a distinctly different design from Intel's tri-gate.

* Not a spokesperson for Intel Corp. *

pm, thanks for your insight, but could you please elaborate on the difference between "dual-gate" FinFET and "tri-gate" one. From my understanding the "gate" here refers to surfaces of the fin, so instead of one planar surface, you have three, two sidewalls of the fin and the top of it. To me it seems that every rectangular fin sticking out has three surfaces, including this one from TSMC (IEDM 2010)

Is this just an accounting issue where Intel is counting the top of the fin as a gate, getting a "tri gate" while everyone else is only counting the sides of the fin, getting a "dual gate?" Or are there fundamental differences. I can imagine that if the fin looked like a triangular roof sticking out, that would be "dual gate".

 

[Arkadrel]

I still cant help but think that the biggest impact this will have will be for Intels Atom line-up (not its Ivy bridges ect).

A 22nm Tri-gate'd Atom is the goal for Intel, the faster (its launched) the better.

** Bonus points if Intel can get a small Atom processor out that can go into tablets, run windows 7, right now before Windows 8 is even launched for arm processors.

** Double bonus points, if its performance/watt kicks the shyt out of the ARM tablets.

 

[TonyB]

Who else here believes Intel is part of a government conspiracy which has aquired alien technology and is now currently exploiting it in an attempt to "launder" the technology as an indigenous Earth base advancement?