David Schor: Intel 10nm in big problems

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
22,701
303
126
It looks like I'm getting a 7nm iPhone this year...

It's amazing just how much the computing landscape has changed in the last decade. Ten years ago if anyone had told me I'd be getting an iPhone with an Apple designed SoC that was fabbed with a more advanced process than Intel's best (high volume) process, I'd have recommended that person be committed to the insane asylum.
 

dark zero

Platinum Member
Jun 2, 2015
2,534
97
91
Seems that they are using EUV... is the real deal... they might win this fight.... and Intel and Samsung will be on problems.
 

Gideon

Senior member
Nov 27, 2007
804
1,191
136
Seems that they are using EUV... is the real deal... they might win this fight.... and Intel and Samsung will be on problems.
Samsung is using EUV as well. Rather GlobalFoundries is on the backfoot
 

ksec

Senior member
Mar 5, 2010
411
101
116
To be fair, WWDC is not really the Apple event where very new hardware is released. But yeah, Apple has a pretty good idea of the hardware they are able to deliver in September by the time of WWDC. We will just not get to know by then.

MacRumors has some experience in following Apple and made a prediction on the sort of things they will release at WWDC: https://www.macrumors.com/roundup/wwdc/

btw, a nice overview of the different actual sizes of the different company's lithography processes ought to be nice in the discussion.
For the past 2 years, Apple has been releasing iPad ( Pro ) in WWDC. And the SoC were using the latest node, to ensure everything works smoothly before bumping out iPhone SoC in volume.
 

jpiniero

Diamond Member
Oct 1, 2010
7,799
1,114
126

Thunder 57

Golden Member
Aug 19, 2007
1,408
1,183
136
Going to be more than that now. Broadwell-Y shipped in Q3 14.

Def seems like good odds now you will see Matisse before a 10 nm desktop part.
I have faith in 7nm because IBM, but it is still GloFo we're talking about. I wouldn't expect Zen 2 until 2H 2019.
 

ksec

Senior member
Mar 5, 2010
411
101
116
No wonder why Apple is switching away from Intel on Mac. Or at least the rumours said.

So Francois Piednoel is wrong, again.

We will have to wait TSMC to actually deliver 7nm, assuming iPhone new SoC will be on 7nm.

Then we can officially confirm, Intel's manufacturing lead is finally over.
 

asendra

Member
Nov 4, 2012
156
12
81
No wonder why Apple is switching away from Intel on Mac. Or at least the rumours said.

So Francois Piednoel is wrong, again.

We will have to wait TSMC to actually deliver 7nm, assuming iPhone new SoC will be on 7nm.

Then we can officially confirm, Intel's manufacturing lead is finally over.
Even earlier. New iPad Pros with A11X on 7nm in WWDC June, most likely.
 

LTC8K6

Lifer
Mar 10, 2004
28,523
1,568
126
Well, those 14nm chips still seem to be holding the high ground, so Intel still has time to get 10nm going.
 

coercitiv

Diamond Member
Jan 24, 2014
3,726
3,509
136
Well, those 14nm chips still seem to be holding the high ground, so Intel still has time to get 10nm going.
Depends on GloFo 7nm timing and quality. Current 14nm is holding the door with a sizeable advantage in maturity, 10nm won't have this luxury.
 

LTC8K6

Lifer
Mar 10, 2004
28,523
1,568
126
Depends on GloFo 7nm timing and quality. Current 14nm is holding the door with a sizeable advantage in maturity, 10nm won't have this luxury.
If we see the magical 8 core Coffee Lake chip, will it likely rule the desktop roost?

Intel's 10nm may be more mature than we would think, if they have gone right to 10nm+?
Is Intel's 10nm still more like everyone else's 7nm?
 

ksec

Senior member
Mar 5, 2010
411
101
116
Even earlier. New iPad Pros with A11X on 7nm in WWDC June, most likely.
Yes and I have said it in my earlier reply. I am just assuming iPad Pro doesn't count as "volume" shipment. iPhone surely does.

I am not expecting any IPC improvement from Apple. Even if they just bump up the core count and some Mhz will be good enough for me*. But if they do, holy cow we are in a year where Intel's dominance fail in both IPC and Manufacturing.

*I am guessing it will be same core, but 4+4 Core and Double the Graphics Core, assuming everything else stay the same, its die size should be noticeably smaller in 7nm. 10nm to 7nm provide ~40% area reduction. The final die size of A12 may likely be in the range of 6x mm. Much smaller then the 84mm. To absorb some of the cost in 7nm production.
 

jpiniero

Diamond Member
Oct 1, 2010
7,799
1,114
126
Yes and I have said it in my earlier reply. I am just assuming iPad Pro doesn't count as "volume" shipment. iPhone surely does.
Apple did sell 13M iPads in 4Q. Can't say how many of those are iPad Pro versus the cheaper models though. It'd have to be at HVM to produce an iPad Pro.

Depends on GloFo 7nm timing and quality. Current 14nm is holding the door with a sizeable advantage in maturity, 10nm won't have this luxury.
Matisse is built to yield. They could sell chips where a good portion of the die is defective and be okay with it.
 

Eug

Lifer
Mar 11, 2000
22,701
303
126
Apple did sell 13M iPads in 4Q. Can't say how many of those are iPad Pro versus the cheaper models though. It'd have to be at HVM to produce an iPad Pro.
Dunno about 2017, but in 2016, the iPad Pro was less than 1/3rd of all iPad sales.

But that would still mean volumes of several million per quarter. What is the definition of "high volume"?
 

DisEnchantment

Senior member
Mar 3, 2017
527
929
106
  • Like
Reactions: lightmanek

tamz_msc

Platinum Member
Jan 5, 2017
2,554
2,245
106
Charlie recently tweeted that alongside TSMC ramping up 7nm, he's also seen working chips on Samsung's 7nm and GloFo being on track as well. By this time next year we might have all four leading edge foundries attaining parity, provided Intel can salvage its 10nm.
 

dark zero

Platinum Member
Jun 2, 2015
2,534
97
91
If we see the magical 8 core Coffee Lake chip, will it likely rule the desktop roost?

Intel's 10nm may be more mature than we would think, if they have gone right to 10nm+?
Is Intel's 10nm still more like everyone else's 7nm?
Hardly.. EUV cuts that advantage Intel has....
 

dark zero

Platinum Member
Jun 2, 2015
2,534
97
91
Yes and I have said it in my earlier reply. I am just assuming iPad Pro doesn't count as "volume" shipment. iPhone surely does.

I am not expecting any IPC improvement from Apple. Even if they just bump up the core count and some Mhz will be good enough for me*. But if they do, holy cow we are in a year where Intel's dominance fail in both IPC and Manufacturing.

*I am guessing it will be same core, but 4+4 Core and Double the Graphics Core, assuming everything else stay the same, its die size should be noticeably smaller in 7nm. 10nm to 7nm provide ~40% area reduction. The final die size of A12 may likely be in the range of 6x mm. Much smaller then the 84mm. To absorb some of the cost in 7nm production.
On the iPads I can see what you said.
On iPhones I see 2 small cores and 1 extra GPU core inside the chip.
 
May 11, 2008
18,310
828
126
Has anyone ever mentioned what exactly is going wrong with the process ?

Maybe they are going for something new.
I remember that long ago Intel was the first with strained silicon in their 90nm process.
Back to memory lane :
https://www.eetimes.com/document.asp?doc_id=1145018

Meanwhile, Intel's decision may boost the prospects of MIT spin-off AmberWave Technology (Nashua, N.H.). The 40-person company has developed a form of strained silicon that it has licensed to Advanced Micro Devices and to new AMD technology and foundry partner UMC Corp., among others.

Until last week, Intel had remained quiet about strained silicon, breathing not a word of interest at the various academic conferences, where the technology has become a mainstream research topic. As recently as June, at the 2002 Symposium on VLSI Technology in Honolulu, Intel managers adopted a skeptical pose, questioning whether strained silicon as IBM and others had described it would be worth the additional costs (see June 17, page 1).

But behind that curtain of public disdain, Intel was moving strained silicon technology into test manufacturing. In March it announced that test SRAM chips were being made at 90-nm design rules at its development fab in Hillsboro, Ore. What it failed to mention was that strained silicon was part of the process flow.
What i posted in the other thread about Molybdenum-Disulfide , is also Stanford and MIT collaboration.
 

ASK THE COMMUNITY