Intel Has 14 nm Test Circuits In The Lab, Limited Teaser Info Released

[qubit]

Nordic Hardware has scored an exclusive interview with Pat Bliemer, Managing Director of Intel Northern Europe to discuss the technology following on from the 22 nm one used in the upcoming Ivy Bridge processors. Unfortunately, Bliemer was light on the technical details of this technology and didn't say when it would see the light of day, except to say that it will make fuller use of the Tri-Gate tech being used in the Ivy Bridge processors and that test circuits are running.

We need to keep going and you can trust me that in our labs we actually have the next generation after 22nm running, so we need to keep going.[...]I cannot really disclose more about that other than that in a laboratory-environment, absolutely we do have the path, our engineers do have the path to actually go and produce 14nm products. There are many variables that you can play with of course it is not the right name for it and the engineers would not like it when I say play, that you can influence to actually go and stay to that model. And I think the breakthrough we had now with the 3D metal gates, just the design of the gate will actually allow for much more efficient thermals and power.

Back in July, we looked even further ahead and reported that Intel aims to have 10 nm-based processors by 2018 and that the 14 nm tech is due for release around 2014. Watch this space.



View at TechPowerUp Main Site

 

[pantherx12]

How in the hell are AMD ( or anyone else) going to be able to compete with this?

lol

 

[qubit]

They're dead in the water.

 

[v12dock]

How in the hell are AMD ( or anyone else) going to be able to compete with this?

lol

The APU

 

[pantherx12]

The APU

I don't think APUs can save AMD from this ( I think this is why they're backing out the high end market)

14nm = potentially packing double he amount of transistors that AMD can in the same amount of space ( based on AMDs 2014 road map)

3d transistors are much more power effciant they have much lower voltage leak and can because of this can do higher clocks at lower voltage.

That means over-clocking monsters as heat shouldn't be a problem.*

* There will still be heat, but there will be far less of it

 

[qubit]

Unfortunately for AMD, from the Ivy Bridge generation, the IGP unit is going to be very competitive with theirs in terms of raw 3D performance and DX11 support and will actually make a decent low end "graphics card". No, Intel are totally driving this market in a one horse race.

 

[Solaris17]

oh man AMD is F#$%d

 

[AphexDreamer]

How in the hell are AMD ( or anyone else) going to be able to compete with this?

lol

By not and focusing on Hybrid GPU's that scale well in other ways?

 

[pantherx12]

AMD needs to come up with their own version of a 3d transistors or pay a licensing fee or something to stay in the game, that can kick global foundries up the arse so they can keep at at the NM race.

That or make a breakthrough in optic computing or graphene chips lol

By not and focusing on Hybrid GPU's that scale well in other ways?

With this much of an architectural edge even intels crappy GPU technology would be able to match AMDs if not beat it. ( so intel will be able to beat AMD at their own game)


No wonder AMD are going mobile

 

[AphexDreamer]

AMD needs to come up with their own version of a 3d transistors or pay a licensing fee or something to stay in the game, that can kick global foundries up the arse so they can keep at at the NM race.

That or make a breakthrough in optic computing or graphene chips lol



With this much of an architectural edge even intels crappy GPU technology would be able to match AMDs if not beat it. ( so intel will be able to beat AMD at their own game)


No wonder AMD are going mobile

HYBRID CPU's!!!!11!~!Q@!!!!

 

[Moose]

I doubt intel will ever produce decent graphics at least AMD has ATi.

Also http://www.techpowerup.com/forums/showthread.php?t=155920 I don't think AMD are planning on trying.

 

[bear jesus]

AMD needs to come up with their own version of a 3d transistors or pay a licensing fee or something to stay in the game, that can kick global foundries up the arse so they can keep at at the NM race.

As a partner in the Joint Development Alliance (JDA) centered around IBM, GlobalFoundries has access to over a decade of research in finFET technology and can choose to use it when it may be needed. The JDA’s collaborative decision to continue with 2D planer structures in 20nm was based on the power, performance, and cost parameters that drive the specifications for the global industry, from high performance desktop computing to low power mobile applications.

And a little info on FinFETs.

 

[ensabrenoir]

 

[ayban]

looking forward to Intel Radeon Graphics or maybe Intel Larabee HD8xxx by 2014.

 

[function69]

Wake me up when it's 2nm

 

[SteelSix]

Damn. Almost scary.

At this pace, Skynet will be online soon.

 

[NC37]

AMD needs to come up with their own version of a 3d transistors or pay a licensing fee or something to stay in the game, that can kick global foundries up the arse so they can keep at at the NM race.

That or make a breakthrough in optic computing or graphene chips lol



With this much of an architectural edge even intels crappy GPU technology would be able to match AMDs if not beat it. ( so intel will be able to beat AMD at their own game)


No wonder AMD are going mobile

Making a decent GPU isn't all performance. Even if they get something able to match an APU, they do not have mature drivers. It would take total refocusing from Intel for them to turn their GPUs around on every front.

Plus, I'll believe it when I see it. Intels mindset is to throw enough x86 at it to surpass the weakness. Real GPU cores are much more advanced. Until Intel changes that mindset, it is foolish to think they'll produce a competitive GPU.

 

[Damn_Smooth]

The GPU argument is fun and all, but it doesn't matter when AMD can't produce a CPU powerful enough to push it's graphics. With the 7000 series rumored to be twice as powerful, we have officially passed the point where their CPUs can keep up. I will be running discreet in my main system until they no longer let me. I don't want to speak for anyone else, but I think most of you will be doing the same.

 

[EpicShweetness]

The GPU argument is fun and all, but it doesn't matter when AMD can't produce a CPU powerful enough to push it's graphics. With the 7000 series rumored to be twice as powerful, we have officially passed the point where their CPUs can keep up. I will be running discreet in my main system until they no longer let me. I don't want to speak for anyone else, but I think most of you will be doing the same.

Very true the current line of "high end" CPU's from AMD can not take advantage of all the raw power there GPU's have. However there is one problem. The APU's combination of a decent CPU and GPU is for HTPC or the everyday user, they don't need all the x86 power of an Intel CPU, there CPU's do everyday task's just fine. Plus alot of things out there are being offloaded onto the GPU.

 

[Completely Bonkers]

AMD is in trouble. I'm sure there are one or two decent engineers there... but with a little headhunting and for (relatively) small change, Intel can strike AMD in the heart now by taking out just a few technical engineers and giving them jobs at Intel. Never mind any gardening leave... AMD is listing and it will only take a small nudge to sink it for years.


PS. This PR is a bit of scare mongering too I think. I doubt Intel will go straight to 14nm. It is two jumps from 22nm. I expect 16nm or 18nm first.

 

[Lipton]

PS. This PR is a bit of scare mongering too I think. I doubt Intel will go straight to 14nm. It is two jumps from 22nm. I expect 16nm or 18nm first.

They do actually plan on going 14 nm after 22 nm. See this link for reference: http://www.intel.com/content/www/us/en/silicon-innovations/intel-tick-tock-model-general.html

Intel with this 'news' to AMD: Gotta love tech though, eh?

 

[Completely Bonkers]

I wonder if the strategic roadmap and actual practice will follow. There must be some interesting new silicon technology to jump such a large factor 22nm to 14nm. Perhaps technically the lithography is at 14nm but the "spacing" of components will be wider to deal with current leakage etc. I guess we will not know until they launch and we do a transistor count per mm2. 22nm is truly amazing. 14nm is incredible! To think I am still running a 130nm Northwood!

 

[de.das.dude]

i dont think intel will be able to compare to AMDs IGPs

 

[Trackr]

So..

IB (22nm) - 2012
Haswell (14nm) - 2014
Broadwell (10nm) - 2018

We're going to be stuck with Ivy Bridge for two whole years?

We're going to be stuck with Haswell for FOUR YEARS?!

 

[Lipton]

So..

IB (22nm) - 2012
Haswell (14nm) - 2014
Broadwell (10nm) - 2018

We're going to be stuck with Ivy Bridge for two whole years?

We're going to be stuck with Haswell for FOUR YEARS?!

It's not quite that bad.

Sandy Bridge (32 nm, 2011) -> Ivy Bridge (22 nm, 2012).
Haswell (22 nm, 2013) -> Broadwell (14 nm, 2014).
Skylake (14 nm, 2015) -> Skymont (10 nm, 2016).

Intel source [PDF]. This, of course, may be subject to change with revised roadmaps.