or Connect
AppleInsider › Forums › General › General Discussion › TSMC forecasts near 100% share of 28nm chip market in 2013, may signal massive Apple orders
New Posts  All Forums:Forum Nav:

TSMC forecasts near 100% share of 28nm chip market in 2013, may signal massive Apple orders - Page 2

post #41 of 54
Quote:
Originally Posted by iSteelers View Post


Actually the fx8350 out performs the i7 in certain tasks.

That's a pointless statement.

A Hyundai Elantra outperforms a Ferrari in certain tasks, too.
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
post #42 of 54
Quote:
Originally Posted by Mechanic View Post

I think one of the reason adnandtech gave was because arm is intels atom series chip's direct competition.  I dont think they want to help the demise of there own product more than has already happened.

 

There was a rumor going around last december that Apple was in talks with Intel.  I believe those rumors as I'm sure Apple talks with everybody.

 

At the Q1 2012 conference call in spring, Paul Otellini made the comment

 

"

Paul S. Otellini - Chief Executive Officer, President, Director and Member of Executive Committee

 

Are there opportunities out in time to take advantage of the lead that we're building in areas like foundry? Yes, perhaps, and I've said that before. And you've seen some small announcements to that effect, where we've signed up some companies for some foundry activity over the next several years. I would look at those, and I would ask you to look at those as being learning experiences for us. A lot of the work we're doing here is to build the libraries, the tools that allow us as a company and our designers of the company to be able to use quick time to market, quick derivative capabilities, modularity for our SoC businesses going forward. And the best way to test those is to have some third-party customers to really validate how good they are, how -- or where they need some work. So it's a learning expense for us. It is for profit, it is not done strictly for -- just to get to know how to do this. And in terms of where it goes long term, I'll leave that point open.

...

Christopher B. Danely - - JP Morgan Chase & Co, Research Division

 

Actually, Paul, just a follow up on Ross' question. I mean is there anything that would prevent you from getting Apple's processor business for the foundry or QualComm's business? I mean, it would be kind of competing against the Atom. But is that theoretically possible? Are you opposed to anything like that?

 

Paul S. Otellini - Chief Executive Officer, President, Director and Member of Executive Committee

 

Well, anything is theoretically possible. The things that would stand in the way with that, both of those, would be the right commercial agreement. I have to say that from a -- from the kind of taste it would leave my mouth, the Apple win would be a little bit -- a lot more attractive than the QualComm win."

 

http://seekingalpha.com/article/505221-intel-s-ceo-discusses-q1-2012-results-earnings-call-transcript

 

(Login required so I included more of the text around the interesting bits.)

 

That's not "no indication".  That's the CEO of Intel going nudge, nudge, wink, wink...Cook just needs to offer me the right kind of deal in response to a direct question.  As I stated earlier I believe an iPad design win for Atom would make it a no brainer for Intel to say yes.

 

Whether that's something Apple is willing to do it depends on whether Intel's mobile roadmap looks as good as Intel's desktop roadmap in 2005.  In many ways Apple faces the same issues in terms of availability of CPUs in the volumes and timeliness they desire.  In this case it's because of a desire to move away from Samsung but the remaining players have all experienced production issues (TSMC and GF) to make a transition shaky.

 

Intel is already a node and half ahead and with 3-D at full rate 22nm production and in fully funded transition to 14nm and 450mm wafers.  450mm wafer transition is going to be hugely expensive and the only guys left will be Intel, Samsung, TSMC and Global Foundries.  The D1X fab in Oregon, Fab 42 in Arizona and Fab 24 in Ireland are prepping to start trial production of 14nm by then end of this year.  Those fabs are all paid for and built...unlike say, UMC fabs.

 

If Samsung is out and TSMC and GF often shaky then Intel is a safe harbor.  Transition to Intel mobile chips in the 2016 timeframe seems to me not to be the outlandish choice but something of the obvious one.  2015 would be amusing simply because it would be a decade after then previous Intel transition. 

post #43 of 54
Quote:
Originally Posted by SockRolid View Post

Re: "may signal massive Apple orders"

Ya think?

I would say the no duh at this point is that analysts are trying to pump TSMC stock with the association to Apple.

It's possible Apple could have set up a deal with them, possible there is someone else out there that we don't know about and that is who will be doing Apple's work, or even possible Apple will go with something totally different chip wise. Who knows at this one.

A non tech's thoughts on Apple stuff 

(She's family so I'm a little biased)

Reply

A non tech's thoughts on Apple stuff 

(She's family so I'm a little biased)

Reply
post #44 of 54
Quote:
Originally Posted by drblank View Post

I read somewhere that TSMC and Apple were working on 20nm processors.  Here's the link.

http://www.digitimes.com/news/a20130116VL201.html

Thanks for the laugh. I needed that

A non tech's thoughts on Apple stuff 

(She's family so I'm a little biased)

Reply

A non tech's thoughts on Apple stuff 

(She's family so I'm a little biased)

Reply
post #45 of 54
Quote:
Originally Posted by nht View Post

Intel is already a node and half ahead and with 3-D at full rate 22nm production and in fully funded transition to 14nm and 450mm wafers.  450mm wafer transition is going to be hugely expensive and the only guys left will be Intel, Samsung, TSMC and Global Foundries.  The D1X fab in Oregon, Fab 42 in Arizona and Fab 24 in Ireland are prepping to start trial production of 14nm by then end of this year.  Those fabs are all paid for and built...unlike say, UMC fabs.

This amazes me every time I see it. I remember when it was argued that we'd never see 200 nm. I wonder what the real limit will actually be. We have to get there someday - when lines are just a couple of atoms wide.
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
post #46 of 54

Interesting that so many people consider this report more credible than the others (like the ones suggesting reduced orders)?

post #47 of 54
Quote:
Originally Posted by nht View Post

/shrug

Yes, it's cheaper and priced against the Core i5 market but its still considered their flagship CPU.  The contention was that Intel can't scale vs AMD which is a weird assertion given the TDP and IPC advantage Intel enjoys.

Bulldozer was a disappointment and Piledriver isn't that much better.  Piledriver is what Bulldozer should have been.  The FX line USED to be top tier in terms of price and performance...not a budget part that competes against Intel's second string.

In terms of Intel vs ARM, Intel appears to me to be driving power usage down faster than ARM is driving performance up.  10W Haswell is going to have very impressive performance-per-watt numbers.   Enough so that Tom's reports:

"Intel is picking the easiest competitor to beat (Tegra 3 running under Windows RT on a Microsoft Surface), but our own preliminary estimates suggest the 32 nm Atom is going to be roughly equivalent to Qualcomm's 28 nm APQ8060A in the ATIV Tab, and more efficient than the 32 nm Exynos 5 Dual in the Chromebook Series 3 XE303C12.
...
Before we had actual Atom-based hardware running in our labs, it was easy to accept that ARM was more efficient than x86. The Cortex-A9 core is slower than Atom, so it should be more efficient. Then, when we started seeing the -A15's performance numbers and how much faster that architecture is, x86 appeared to be doomed. But nobody really stopped to ask whether -A15's power consumption was still as good, nor had they sat down to run the numbers.  Once we started looking at battery life for a given capacity, we began to think about power use in terms of actual watts consumed during a particular task."

http://www.tomshardware.com/reviews/atom-z2760-power-consumption-arm,3387-5.html

Anand writes something similar:

"We already know that Atom is faster than Krait, but from a power standpoint the two SoCs are extremely competitive.
...
If the previous article was about busting the x86 power myth, one key takeaway here is that Intel's low power SoC designs are headed in the right direction. Atom's power curve looks a lot like Qualcomm's, and I suspect a lot like Apple's. There are performance/power tradeoffs that all three make, but they're all being designed the way they should.
...
I'd always heard about Haswell as the solution to the ARM problem, particularly in reference to the Cortex A15. The data here, particularly on the previous page, helped me understand exactly what that meant. Under a CPU or GPU heavy workload, the Exynos 5 Dual will draw around 4W. Peak TDP however is closer to 8W. If you remember back to IDF, Intel specifically called out 8W as a potential design target for Haswell.
...
The really crazy part is that it's not too absurd to think about being able to get a Core based SoC into a large smartphone as early as 14nm, and definitely by 10nm (~2017) should the need arise"

http://www.anandtech.com/show/6536/arm-vs-x86-the-real-showdown/14

If Intel isn't leading today, it's come a helluva long way and looks like it could lead in the future. Anand seems bullish on the idea.

If I can get Core i7 performance in an iPad Mini by 2016 I will welcome our new Intel mobile overlords.

There is no doubt that AMD is a fading company. I would hope that Apple doesn't consider them again. The only thing I would question here is 10nm. Despite Intel's assertion that they will be coming up with it when they state, many experts in this business still have doubts about 10nm. Even 14 is more than leading edge. I wonder if we will see 10 when Intel states we will, and I wonder if we will see it at all. It seems to be very aggressive of them to be moving directly to 14 after 22. While extremely small laboratory runs of 14nm chips have come out from a couple of parties, moving to large scale production still isn't assured. While I have no doubt 14nm will be achieved, 10, and even 12nm is being argued as possibly the end run of chip manufacture, until newer technologies are available. I wouldn't be surprised to see 10nm being pushed out by at least a year from their intended schedule.
post #48 of 54
Quote:
Originally Posted by stelligent View Post

Interesting that so many people consider this report more credible than the others (like the ones suggesting reduced orders)?

The reports on reduced orders involve screens, not processors. And that just involved Sharp, who presently, is just one of three screen manufacturers for Apple's mobile devices. The other two being LG and Samsung, neither of which are included in this reduced orders debacle. It's also thought that a reason for Sharps reduced orders is because they have finally solved their problems of IZGO manufacture, and are switching over production lines for that purpose. If it's true that Apple will come out with new lines of phones and tablets by mid year, on a new, and likely plausible twice a year schedule, then production of IZGO screens would need to begin early in the year, as it's probable that, in the beginning at least, Sharp would be the only producer of those screens for Apple.

So that a manufacturer of perhaps less than one third of Apple mobile screen production would be stopping production as the slowest quarters of the year are coming up isn't a sign of slowing demand, other than the seasonal drop that everyone experiences at this time, coupled with the possible changeover of production to a newer screen type.

These rumors are really getting out of hand.
post #49 of 54
Quote:
Originally Posted by melgross View Post


The reports on reduced orders involve screens, not processors. And that just involved Sharp, who presently, is just one of three screen manufacturers for Apple's mobile devices. The other two being LG and Samsung, neither of which are included in this reduced orders debacle. It's also thought that a reason for Sharps reduced orders is because they have finally solved their problems of IZGO manufacture, and are switching over production lines for that purpose. If it's true that Apple will come out with new lines of phones and tablets by mid year, on a new, and likely plausible twice a year schedule, then production of IZGO screens would need to begin early in the year, as it's probable that, in the beginning at least, Sharp would be the only producer of those screens for Apple.

So that a manufacturer of perhaps less than one third of Apple mobile screen production would be stopping production as the slowest quarters of the year are coming up isn't a sign of slowing demand, other than the seasonal drop that everyone experiences at this time, coupled with the possible changeover of production to a newer screen type.

These rumors are really getting out of hand.


If I am not mistaken, the silly part of the screen reduction rumor is the # of the units (65M?).


Regardless, I think all of these rumors are not worth losing (or gaining) sleep over. Apple's earnings will be announced in 3 days.

post #50 of 54
Quote:
Originally Posted by melgross View Post

There is no doubt that AMD is a fading company. I would hope that Apple doesn't consider them again. The only thing I would question here is 10nm. Despite Intel's assertion that they will be coming up with it when they state, many experts in this business still have doubts about 10nm. Even 14 is more than leading edge. I wonder if we will see 10 when Intel states we will, and I wonder if we will see it at all. It seems to be very aggressive of them to be moving directly to 14 after 22. While extremely small laboratory runs of 14nm chips have come out from a couple of parties, moving to large scale production still isn't assured. While I have no doubt 14nm will be achieved, 10, and even 12nm is being argued as possibly the end run of chip manufacture, until newer technologies are available. I wouldn't be surprised to see 10nm being pushed out by at least a year from their intended schedule.

 

Intel seems to have a non-EUV path to 10mm with quad patterning immersion lithography.  

 

Everything is iffy at these sizes and the end of Moore's Law could be on the horizon.  Then again, that's been said before too and yet, here we are.  I guess when we max out photolithography then technically we're at the end of Moore's Law but the continued exponential increase in computing capability could continue on anyway.

 

Of the companies listed here as the major players (Intel, Samsung, TSMC, GF, etc) the most likely to move forward is Intel followed by Samsung.  TSMC perhaps by default because all these fabless companies need someone to be a foundry besides Samsung or build their own fabs.

 

Apple is one of the few that could afford to do this but Sony is the cautionary tale of building, selling and then repurchasing their fabs due to changes of fortune.  Maybe Toshiba instead since Sony managed to sell their fab for $800M and buy back for $600M when Toshiba elected to go fabless.


Edited by nht - 1/21/13 at 7:37am
post #51 of 54
Quote:
Originally Posted by nht View Post

Intel seems to have a non-EUV path to 10mm with quad patterning immersion lithography.  

Everything is iffy at these sizes and the end of Moore's Law could be on the horizon.  Then again, that's been said before too and yet, here we are.  I guess when we max out photolithography then technically we're at the end of Moore's Law but the continued exponential increase in computing capability could continue on anyway.

Of the companies listed here as the major players (Intel, Samsung, TSMC, GF, etc) the most likely to move forward is Intel followed by Samsung.  TSMC perhaps by default because all these fabless companies need someone to be a foundry besides Samsung or build their own fabs.

Apple is one of the few that could afford to do this but Sony is the cautionary tale of building, selling and then repurchasing their fabs due to changes of fortune.  Maybe Toshiba instead since Sony managed to sell their fab for $800M and buy back for $600M when Toshiba elected to go fabless.

I'm nervous about TSMC as well, because as we know, they always seem to have major problems with every new die shrink.

Truthfully, I'm surprised that Apple and Intel haven't gotten together on this. Even though Intel isn't really interested in being a fab for everyone, this would be a special case. Apple will be needing several hundred million SOC's a year. Even though those are much less expensive that Intel's own "i" series, they will cost the same, or possible slightly more than Intel's own efforts in the mobile space that they have begun embarking upon. While Intel would make less per chip as just a fab, there is still a lot of profit in making several hundred million of what are just two or three different lines at a time, Rather than the dozens and dozens of their current chips that they make.

If Intel has hopes of converting Apple over to x86, I think that would be a real far outside chance. Apple is pretty committed to designing their own offerings, and as long as they see an advantage in doing so, it's very unlikely they would switch over. This isn't like the computer performance race, where software compatibility with Windows, and the lack of a competitive laptop chip were major issues.

I see a lot of benefit to both parties here. I wonder if the issue isn't more of a political nature than a technical or monetary one.
post #52 of 54
Quote:
Originally Posted by pinolo View Post

That is one ugly logo for a company though

 

Quote:
Originally Posted by ksec View Post


Yes. I constantly question why TSMC hasn't updated their Logo. But Since they aren't really a consumer facing company may be they thought it didn't matter.

I haven't read the original info. So some of what this is reported may be not in context of what TSMC was really trying to say. To have 100% 28nm in 2013 would means Samsung wont be doing any 28nm apart from themselves. And GF hasn't secured any 28nm design win.

I still have doubt over Apple moving to TSMC in such a quick pace ( unless it has been planned ahead for years ), otherwise i cant think of any things else that could make TSMC 28nm output nearly triple.


One main function of a logo is to make an impression, TSMC's "Crossword Puzzle" certainly does that.  I'll instantly know it from a distance. And as noted, not looking to be placed in a Trendoid-centric mall between an outlet for Jimmy Choos and a Cheesecake Factory.

Quote:
Originally Posted by Constable Odo View Post

Apple's last gasp at holding decent tablet market share. Samsung is going to flood the market with cheap, low-quality tablets in 2013 in an attempt to put Apple out of the tablet business. Samsung said it has plenty of money to throw away despite losing Apple's component contracts. I'm sure they'll gladly give Android tablet vendors price breaks in order to ramp up production against Apple. Samsung has always tried to stifle the competition and it going to go all-out to cut deeply into iPad market share. I'm glad Apple moved away from Samsung. That company is run by backstabbing crooks who'll do anything in order to steal business away from rivals.

This Samsung is evil meme is getting a little stale.  First, they're pikers in the backstabbing dept. to the MS of the '80's and '90's who really deserved the criticism they got at the time (tho' it's amusing to me when - enfeebled and chastened as they are and led by an ineffective CEO who must have some dirt on Gates to be still the honcho - are still talked about in the same way as when they were the terror of the western business world.  

Second, unlike MS who had a virtual monopoly in their lines of business, every other clone maker out there has the same opportunity to access to all of the same parts bins (outside of Samsung's own parts) as the company, but they're simply out-executing the rest.  That's called being good at what you do.  And one things corporations that reach a certain heft often do is engage in price wars which drive less-capitalized rivals out of business - or at least out of certain businesses. 

If it reaches a certain point, it's called dumping and is illegal.  But charges have to be brought and adjudicated. 

So on this score, HTC, Acer, Asus and Nokia, etc. are taking much more of a beating at Samsung's hands than has Apple to date (tho' doubtless, they have Apple's top tech dog status in their gun sights. 

Third, corporations are inherently amoral.  They use what they have and can think of and get away with to grow and profit.  It's their sole mandate and reason for being. 

 

And Apple is not singularly sainted either.  I heard people cheering here when Apple used its cash hoard to effectively own the world inventory on RAM chips a few years back. 

 

"Brilliant supply chain management" was what it was called here.  I'll bet it was called something else in the HQ's of other tech companies around the world.  So how is that any different or more "moral" than Samsung selling near cost for market advantage?

 

And I personally think the management of Apple is very bright, dedicated and competent at what they do.  But that they haven't replaced that departed "disruptive gene" in their makeup since Steve Jobs died.  It's always an evolving tech world, and corporations can never just keep doing what they've been doing with incremental iterations as the market changes.  And so far, Apple's singular strategy is becoming less competitive in the world of the Android assault where they no longer have "first mover" advantage in any line of business they're in.

Leaving the question, not is Samsung evil, rather, who at Apple is going to demonstrate that the company is still the master of the end-around play that will put the competition back on its heels again for another few years and restores Apple's cachet as both disruptor and true innovator?  Ives refines things.  Cook manages things.  Schiller markets things.  Etc.  But who's going to come up with and shepherd the next new THING from concept to roll-out as Jobs did at least 4 or 5 times?

The question has NOT yet been answered.
 

An iPhone, a Leatherman and thou...  ...life is complete.

Reply

An iPhone, a Leatherman and thou...  ...life is complete.

Reply
post #53 of 54
Quote:
Originally Posted by shompa View Post

The problem is that Apple would need all of TSMC capacity. Apple needs almost 70K wafer starts per month. Thats more then 1 fab line!

BTW. Remember that TSMC had trial runs of A6 in june 2011. Yield problems. Apple's SoC team have tons of work to do. First the A5 32 nm shrink, then the A5X, A6 32nm, A6X 32nm.

I personally don't understand why TSMC would manufacture A6/X. Apple needs to do a new tape out for TSMC. Why not instead do a A7 SoC with PowerVR Rouge? That is what I predict.

We don't know if TSMC has done any R&D for Apple as yet. That's just rumor.
post #54 of 54

You know, something no one has discussed:  If Chang's quote is correct then he expects no 28nm production out of Samsung in 2013.

 

Given that Samsung is rumored to be circulating test versions of the 28nm HKMG A15 'Adonis' processor for their Exynos line I'm wondering if there isn't some mistranslation of his comments.  

 

Either that or TSMC is hiring some ninjas to go sabotage Samsung's 28nm fab...

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: General Discussion
AppleInsider › Forums › General › General Discussion › TSMC forecasts near 100% share of 28nm chip market in 2013, may signal massive Apple orders