• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

[S|A] Broadwell supply a "trickle" until 2015

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
frozentundra123456 said:
I think the bottom line is that all the predictions that TSMC, et. al are going to catch up to intel are just that, predictions. Seems that those who think Intel is going to lose their process advantage are assuming the other suppliers will magically meet every timeline they have predicted without incurring any of the same problems intel has encountered. It is possible they will make a seamless transition exactly on schedule, but it seems like an unlikely scenario to me. OTOH, I think there is a very good possibility that Skylake will be delayed as well. So really, node shrinks are becoming so difficult at this level, it is really difficult to predict what is going to happen.
Click to expand...

i think skylake may have a 2015q4 release, a year after broadwell, but'll probably be staggered in the sku release as is the case here. the overall feel for the summary is that intel has 10nm in solid control, but it took difficulty at 14nm to figure it out.

maybe we'll see iii-v for 10nm after all?
 
KLA-Tencor SEMICON Briefing:

http://files.shareholder.com/downloads/KLAC/3314967070x0x767421/ea503424-fd72-4cc9-8ada-4fb57c7d12cf/S_KLA-520_10_Q4FY14_Exec_Financial_security.pdf

Page 8 has some interesting figures, a push back in expectation for EUV (2013-10nm, 2014-7nm; and 450mm: 2018 -> 2020.

The EUV dilemma continues. Hopefully we can get some more SEMICON info. CNSE/SUNY IT have a slide up on their youtube page from SEMICON today showing their R&D areas. Putting III-V at 10nm, nanowire/gaa at 7nm, and graphene after.

The Analyst Day briefings for Applied Materials are pretty interesting, chart says 10nm = new materials, also there is a chart on electron mobility and new materials, 7.0x for InGaAs compared to silicon.
 
Last edited:
So I guess I shouldn't really have to wait for the MBP refresh this fall? I need a laptop pretty desperately. I found an aging 2008 T61 to replace my Macbook Pro. This screen makes my eyes tear up.
 
jdubs03 said:
KLA-Tencor SEMICON Briefing:

http://files.shareholder.com/downloads/KLAC/3314967070x0x767421/ea503424-fd72-4cc9-8ada-4fb57c7d12cf/S_KLA-520_10_Q4FY14_Exec_Financial_security.pdf
Click to expand...

Quite interesting - thanks for sharing. The historical perspective on EUV is spot on - it just keeps getting pushed out without a clear end in sight. (Notice how it's increased from 2 years out back in 2011 to 4 years out?) Their Foundry FinFET projection of late 2015 is also rather intriguing, in large part for the simple fact that it pushed out a full year compared to the 2013 perspective.

The III-V possibility for Intel's 10nm sounds about right - they have been introducing major process enhancements every other node since strained silicon at 90nm. Though transitioning to III-V materials would put any of those prior enhancements to shame by comparison.
 
mikk said:
Q2 2015 according to Roadmap. Broadwell in the desktop space is independent from Skylake, it will replace Haswell/Haswell refresh non-K.
Click to expand...

Yea, sorry I should have clarified, I expect Skylake-Y/U to be released then.

@khato

no problem. yea from what I've seen Intel wants to keep their process innovation lead intact, so III-V being at 10nm will give them a 1+ node lead.
 
Last edited:
NTMBK said:
Westmere launched in January 2010, Ivy Bridge in May 2012, Broadwell will be December 2014 if we're lucky. The two year cycle has slipped by almost a year.
Click to expand...

Intel confirmed a September-October release for Broadwell, not December. Cannonlake in June 2016, if we're lucky. Half a year in 6 years.
 
jdubs03 said:
i think skylake may have a 2015q4 release, a year after broadwell, but'll probably be staggered in the sku release as is the case here. the overall feel for the summary is that intel has 10nm in solid control, but it took difficulty at 14nm to figure it out.

maybe we'll see iii-v for 10nm after all?
Click to expand...

Skylake will be launched in Q2.

We'll probably get III-V at 10nm, but the same sources also say we'll get Germanium at 10nm, which confuses me.
 
jdubs03 said:
KLA-Tencor SEMICON Briefing:

http://files.shareholder.com/downloads/KLAC/3314967070x0x767421/ea503424-fd72-4cc9-8ada-4fb57c7d12cf/S_KLA-520_10_Q4FY14_Exec_Financial_security.pdf

Page 8 has some interesting figures, a push back in expectation for EUV (2013-10nm, 2014-7nm; and 450mm: 2018 -> 2020.

The EUV dilemma continues. Hopefully we can get some more SEMICON info. CNSE/SUNY IT have a slide up on their youtube page from SEMICON today showing their R&D areas. Putting III-V at 10nm, nanowire/gaa at 7nm, and graphene after.

The Analyst Day briefings for Applied Materials are pretty interesting, chart says 10nm = new materials, also there is a chart on electron mobility and new materials, 7.0x for InGaAs compared to silicon.
Click to expand...

Here's Intel's published roadmap:

http://download.intel.com/newsroom/...esearch_Enabling_Breakthroughs_Technology.pdf

ASML's expectations:

http://www.sokudo.com/event/images/130710/5_ASML.pdf
 
witeken said:
Intel confirmed a September-October release for Broadwell, not December. Cannonlake in June 2016, if we're lucky. Half a year in 6 years.
Click to expand...

Have they? Last I saw was "holiday".

And even if it is October, that's still a 9 month slip since Sandy Bridge.
 
But the roadmap shows the innovations Intel will introduce in the future. Germanium and III-V at 10nm and further in the future, from 5nm on, CNTs, graphene and wires/dots. If you combine that with the other link and AM's roadmap, you get a nice picture of the future.

different-transistor-topologies.jpg
 
Last edited:
NTMBK said:
Have they? Last I saw was "holiday".
Click to expand...

Back to school would be close.

And even if it is October, that's still a 9 month slip since Sandy Bridge.
Click to expand...

Cannonlake might be 6 months since Sandy Bridge. That's 13 months on average.

And don't forget AMD and Nvidia and basically the whole semiconductor industry. They're all doing (much) worse than Intel.
 
witeken said:
Back to school would be close.
Click to expand...

"Back-to-school" and "holiday" are different.

Cannonlake might be 6 months since Sandy Bridge. That's 13 months on average.
Click to expand...

Based on what, precisely? I have seen no roadmaps with Cannonlake on them yet.

And don't forget AMD and Nvidia and basically the whole semiconductor industry. They're all doing (much) worse than Intel.
Click to expand...

Yes, everybody is finding it hard, but we're not discussing them. We're discussing Intel. 😉
 
NTMBK said:
"Back-to-school" and "holiday" are different.
Click to expand...
http://www.reuters.com/article/2014/05/18/us-intel-chips-idUSBREA4H08P20140518

""I can guarantee for holiday, and not at the last second of holiday," Krzanich said in an interview. "Back to school - that's a tight one. Back to school you have to really have it on-shelf in July, August. That's going to be tough."

I don't consider October and certainly not December "tight".

Based on what, precisely? I have seen no roadmaps with Cannonlake on them yet.
Click to expand...
1) Skylake's pull-in to Q2'15; 2 years after Broadwell, when you extrapolate Cannonlake is Q2'16

2) 14nm (and 10nm too) is projected to last exactly 2 years.
409118-intel-technology-roadmap.jpg


3) Relentless pursuit of Moore's Law
2013_IM_Krzanich-3.jpg


4) Couldn't be said more explicit:
2013_IM_Krzanich-21.jpg


If they start HVM in Q4'15, there will be products in mid-2016. Don't forget that Broadwell-Y, the first 14nm platform, must be placed in products, so if Cannonlake's first SKUs will be LGA, the TTM (HVM to availability) will be faster as well.

Yes, everybody is finding it hard, but we're not discussing them. We're discussing Intel. 😉
Click to expand...
Intel doesn't see any major difficulties in the coming 10 years.
 
witeken said:
http://www.reuters.com/article/2014/05/18/us-intel-chips-idUSBREA4H08P20140518

""I can guarantee for holiday, and not at the last second of holiday," Krzanich said in an interview. "Back to school - that's a tight one. Back to school you have to really have it on-shelf in July, August. That's going to be tough."
Click to expand...

In other words it's not coming for back to school. Okay. And it will arrive during the holiday period... So anywhere from October to December.

1) Skylake's pull-in to Q2'15; 2 years after Broadwell, when you extrapolate Cannonlake is Q2'16
Click to expand...

So you're making the date up, okay. 😉

2) 14nm (and 10nm too) is projected to last exactly 2 years.
Click to expand...

And 22nm was projected to last exactly 2 years too. Look how that has turned out.

3) Relentless pursuit of Moore's Law
Click to expand...

Yawn, PR fluff. I'm sure TSMC would say that they are "relentlessly pursuing Moore's Law" too.

4) Couldn't be said more explicit:
Click to expand...

It's a good job Intel has never failed to deliver on a similar promise, right? Oh wait, they said 14nm would be ready by the end of 2013.

Look, don't put your faith blindly in everything Intel says. They come up against unforeseen problems just like every other company in the history of the world; they don't have a magic tick-tock clock which guarantees that a process will take exactly two years to bake in the oven before its ready. These estimates all come from a bunch of incredibly clever people who are doing their best to guess at the unknowable, and sometimes estimates just don't work out. It happens. I'm not belittling Intel- they clearly have thousands of extremely intelligent engineers, and they have some flat out fantastic technology. But they're not infallible, because nobody is.
 
NTMBK said:
They come up against unforeseen problems just like every other company in the history of the world;
Click to expand...

This basically sums up your whole post, you're milking Intel's 14nm yield issues to try to make your point. Everything suggests a release of 10nm in mid-2016, and you reply that they could have problems, like 14nm. Quite obvious, isn't it?

But those are just that: unforeseen problems. I don't see why I should take nonexistent problems into account.

SO at 14nm, they ran into yield issues. That doesn't mean the same thing will happen at 10nm. And if they didn't have those issues, like you said, they would have started HVM in Q4'13 and 22nm would have lasted for a shorter time.

Also, from earlier:

ii. The need to use diffracted light created an optical engineering challenge that “exploded” at Intel at 14nm
c. Intel has now developed this this “Inverse Optical Lithography” capability in house, it is a huge competitive advantage
Bottom line: Intel will gain a transistor cost lead for the first time ever at 14nm, and will extend that at 10nm as TSMC and Samsung struggle to solve this Inverse Optical Lithography challenge
Click to expand...

Intel will likely use EUV at 7nm (10nm is also "tight"), so there won't be any lithography problems in the near future. There could be things that result in a delay of 10nm, but I won't update my prediction before I know those. So it's all nice to criticize my forecast, but I bet you can't come up with a better one.
 



The RTS (Release Time Schedule?) only gets shifted out more and more.
In one quarter (=13 weeks) the worst case RTS for quad cores was
delayed by 21 weeks from week 15/2015 (April 2015) to week 36/2015
(September 2015)

Worst case RTS for the first 14nm products is now week 3/2015.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top