TSMC signs chip deal with Apple

[ShintaiDK]

It was expected. TSMC will sooner or later be the sole foundry of the world, besides Intel chips.

Apple has just signed a three year deal with Taiwan Semiconductor Manufacturing Company (TSMC) and its IC partner Global UniChip to produce Apple’s next A-series chipsets with 20, 16 and 10nm process nodes. Neither TSMC nor Global UniChip have commented upon the deal, but they will begin producing the new Apple A8 chip in low volumes as early as July. In December of this year, production capabilities for TSMC will scale up significantly on the 20nm chips, when they begin installing new fabrication equipment. By the first quarter of 2014 they will be capable of processing 50,000 circuit wafers. This equipment can later be upgraded to also produce 16nm chips.

http://www.digitimes.com/news/a20130624PD207.html

 

[MisterMac]

So..... where will qualcomm go next year when it needs volume?

Unless morris has balls the size of the moon - i assume Apple has some serious graft on the terms of volume requirements on bleeding edge.

 

[ShintaiDK]

So..... where will qualcomm go next year when it needs volume?

Unless morris has balls the size of the moon - i assume Apple has some serious graft on the terms of volume requirements on bleeding edge.

TSMC as well. The loser is AMD/nVidia. 20nm GPUs are not coming anytime soon I guess. Apple/Qualcomm bidding war.

 

[mrmt]

So..... where will qualcomm go next year when it needs volume?

Unless morris has balls the size of the moon - i assume Apple has some serious graft on the terms of volume requirements on bleeding edge.

With TSMC. There's a reason for TSMC to go on a CAPEX spree as they are now and this reason is to get Apple *and* to get Qualcomm and the other big volules.

Even if they delay everyone 6 months to accommodate Apple, nobody will leave, as TSMC is essentially with no competition for 20nm.

 

[MisterMac]

...for chips only couldn't a Samsung\Qualcomm alliance make sense?
(assumung samsung ups the ante enough).

 

[NTMBK]

There's always GloFo :awe:

 

[sontin]

TSMC as well. The loser is AMD/nVidia. 20nm GPUs are not coming anytime soon I guess. Apple/Qualcomm bidding war.

TSMC is giving every company access to the first run. There is no real loser. They will have all the same problems.

 

[ShintaiDK]

TSMC is giving every company access to the first run. There is no real loser. They will have all the same problems.

We all know who got most money and willing to pay the most. And its not nVidia or AMD.

 

[krumme]

With TSMC. There's a reason for TSMC to go on a CAPEX spree as they are now and this reason is to get Apple *and* to get Qualcomm and the other big volules.

Even if they delay everyone 6 months to accommodate Apple, nobody will leave, as TSMC is essentially with no competition for 20nm.

The development now have never been so interresting on the market and the foundry business. JHH have noticed and is trying to change NV, - i think he is a good CEO example of what to do under this new situation. He grasps the seriousness and the call for extreme change to survive and thrive.

TSMC have been expanding rapidly for the last years. They just continue this accelerating capex spree.
Foundry business is natural monopoly, that just demands loads of engi hours - not creativity - also for the new design tools for the customers.

The man hours are probably cheaper excactly where TSMC is situated compared to Intel or GF. I think this is a main reason for the change and speed of it. They can simply make it cheaper.
Add the huge ARM ecosystem feeding it, including the major player Qualcomm.
Add NV and AMD can go nowhere else else as of now
Add Samsung and Apple favor TSMC to Intel for strategic reasons

I think TSMC stands to win this race from Intel. Intel can continue to use its profitable servermarket as a leverage, and the integration of process and design, but its not enough to compete against everyone and his brother.
They have labelled themselves, producer of cpus. A stupid image, because they bought their own cpus. CPU that had a purpose. Now the purpose and the usage pattern is changing.

There will be no profitable scrapes left for the small foundries like GF. Nothing. The market is consolidating at an accelerating speed.

In the last interview with Otellini he was saying Apple all the time. He should have done that 6 years ago. Now its different and way to late. Now its Qualcomm and Samsung.

Intel needs to change their foundry business model. Fundamentally. Now. Before its to late.

 

[sontin]

TSMC's business modell is to secure all designs. They don't care for a short time win if this could mean others companies going to the competition.

 

[SiliconWars]

It was expected. TSMC will sooner or later be the sole foundry of the world, besides Intel chips.



http://www.digitimes.com/news/a20130624PD207.html

I guess you missed this and this while reading digitimes?

 

[ShintaiDK]

I guess you missed this and this while reading digitimes?

I didnt miss it. Only GloFo misses everything. Lots of talk, not so much products from their side. So I consider that flop of a company irrelevant.

 

[SiliconWars]

I didnt miss it. Only GloFo misses everything. Lots of talk, not so much products from their side. So I consider that flop of a company irrelevant.

And let me guess, they also fabricated their 31% increase in revenue last year? :whiste:

Oh btw, wasn't Intel going to be fabbing for Apple? I wonder what happened to that...

 

[nehalem256]

And let me guess, they also fabricated their 31% increase in revenue last year? :whiste:

WSA from AMD?

 

[ShintaiDK]

And let me guess, they also fabricated their 31% increase in revenue last year? :whiste:

Oh btw, wasn't Intel going to be fabbing for Apple? I wonder what happened to that...

From 3.5B to 4.5B, yes thats fantastic until you look on the competition. TSMC grew from 14.6B to 17B.

TSMC is on 20nm, when GloFo is on 28nm. A node TSMC already milked leaving discount products to GloFo.

 

[Arachnotronic]

From 3.5B to 4.5B, yes thats fantastic until you look on the competition. TSMC grew from 14.6B to 17B.

TSMC is on 20nm, when GloFo is on 28nm. A node TSMC already milked leaving discount products to GloFo.

TSMC isn't on 20nm yet...volume production happens in 1H 2014.

 

[SiliconWars]

From 3.5B to 4.5B, yes thats fantastic until you look on the competition. TSMC grew from 14.6B to 17B.

So TSMC had half the growth of GF, right.

TSMC is on 20nm, when GloFo is on 28nm. A node TSMC already milked leaving discount products to GloFo.

TSMC is nowhere near production 20nm.

 

[SiliconWars]

WSA from AMD?

Revenues from AMD would have been down last year compared to previous years.

 

[akugami]

I think the biggest entities this affects is Samsung, AMD and nVidia.

Samsung - Obviously this takes billions off the table. Even though Samsung's profit margin may not be much producing Apple's "A" series of ARM processors, that's still a huge blow. These things tend to have a trickle down effect that could affect more than just the dollars Samsung was getting from Apple. Things like higher procurement cost due to lower purchasing volumes can't be ignored.

AMD & Nvidia - Both could lose out because Apple is going to be "stealing" some of their capacity. Especially in an early node. I'm wondering if this will cause a delay in getting new products out. Perhaps extremely limited "new" video cards as both AMD and nVidia wait for chips. Maybe we'll see the return of paper launches as the norm...

 

[BallaTheFeared]

10nm in three years, tell me I can't read!

 

[SiliconWars]

AMD won't be releasing 20nm graphics cards until the end of 2014 so it barely affects them. It may affect Nvidia if they are truly going to release 20nm Maxwell in Q2 2014, which for me is starting to look unlikely.

 

[USER8000]

AMD & Nvidia - Both could lose out because Apple is going to be "stealing" some of their capacity. Especially in an early node. I'm wondering if this will cause a delay in getting new products out. Perhaps extremely limited "new" video cards as both AMD and nVidia wait for chips. Maybe we'll see the return of paper launches as the norm...

AMD and Nvidia usually release their halo high end GPUs first for the consumer market,and release the larger volume parts months later.

If anything you might find,the larger volume parts launching first since they use smaller GPUs,and higher end parts either launching later or at a massive price premium if launching earlier.

 

[Khato]

Quite interesting, especially the fact that Apple is supposedly breaking with tradition and manufacturing their latest design on the leading edge of a new process node. (They were previously quite conservative in that regard, note that they're still using Samsung's 32nm process node for their current products.)

Wonder how long this deal will last/what kind of penalties Apple forced into the contract if TSMC has their typical mis-steps on the new process and can't deliver to Apple's time table? I'd expect that for 20nm they must have been able to present Apple with adequate data that they were going to make their current schedule... which is, by the way, 'only' two to four quarters behind their schedule 2 years ago of beginning 20nm volume production in the second half of 2012 - http://www.eetimes.com/electronics-news/4215986/Design-starts-triple-for-TSMC-at-28-nm

 

[Arachnotronic]

Quite interesting, especially the fact that Apple is supposedly breaking with tradition and manufacturing their latest design on the leading edge of a new process node. (They were previously quite conservative in that regard, note that they're still using Samsung's 32nm process node for their current products.)

Wonder how long this deal will last/what kind of penalties Apple forced into the contract if TSMC has their typical mis-steps on the new process and can't deliver to Apple's time table? I'd expect that for 20nm they must have been able to present Apple with adequate data that they were going to make their current schedule... which is, by the way, 'only' two to four quarters behind their schedule 2 years ago of beginning 20nm volume production in the second half of 2012 - http://www.eetimes.com/electronics-news/4215986/Design-starts-triple-for-TSMC-at-28-nm

20nm does not begin until Q2 2014 in any real volume, as TSMC's latest earnings call stated that 20nm does not become ~2% of the firm's revenues until Q2 2014.

I would think Apple business would mean far more than 2%, so I would take all of this with a grain of salt.

 

[Khato]

20nm does not begin until Q2 2014 in any real volume, as TSMC's latest earnings call stated that 20nm does not become ~2% of the firm's revenues until Q2 2014.

I would think Apple business would mean far more than 2%, so I would take all of this with a grain of salt.

Well, that further tidbit does raise another point of interest - it's likely that the 20nm A8 that the article is talking about won't be in actual products 'til basically Q4 of 2014. (Refresh later this year should be using A7 if they continue their current numbering scheme.) So either TSMC isn't going to have 20nm products shipping until after Intel likely releases 14nm Broadwell or Apple is still giving themselves a bit of buffer between planned process availability for mass production and when they actually need it to work.