Semiwiki on LinkedIn Semiwiki on Facebook Semiwiki on Twitter Register / Log in 
Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/can-apple-save-tsmc.6595/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Can Apple Save TSMC?

C

CharlieD

New member
I would be interested in opinions on this article:

Even Apple Can't Save TSMC

"Even Apple can’t save TSMC, said Maybank‘s Warren Lau, adding:
TSMC now expects FY15 sales to be up slightly less than 10%, lower than previous guidance of well over 10%. 2H sales are now expected to contract 3-5% HoH compared to previous guidance of better 2H than 1H. These imply a third consecutive misguidance in FY15.

We expect the current inventory digestion to be largely over in 4Q15 but a recovery in 1Q16/FY16 may not necessary materialize as demand remains uncertain. Besides, TSMC is losing QCOM’s S820 orders (high-end AP to Samsung), which typically ramp in 4Q/1Q to support new Android premium models.

Also, TSMC remains the second source for Apple’s A9 which is subjected to higher risk should demand falter. Maintain SELL."

​

Sometimes I wonder if our industry can keep up the fast pace we have set and the resulting CAPEX. The same goes for Samsung I guess. Intel will always have the profitable server business but the foundries are fighting for the hyper growth SoC business (clients) which could easily be hyper decelerating the next few years. Opinions?
 
Hi Charlie,

Apple is the one driving the process nodes into a brick wall. 20nm in 2014, 14nm in 2015, 16FFT in 2016, and 10nm in 2017. That is a very fast and very expensive pace for sure. The good news is that Apple has the money to spend on this. The bad news is that if you are not the foundry of choice you may not have the money to spend on this. Picking Samsung for 28nm, TSMC for 20nm, Samsung for 14nm, and TSMC for 16nmFFT (a different foundry for every other node) could be the strategy to keep two competing foundries in the game. If Apple sticks with one foundry for two nodes in a row the other foundry may be in CAPEX trouble. Apple certainly is the foundry king maker now.

D.A.N.
 
Maybank has always had very negitive views on TMSC. Considering they had a sell rating throughout 2014, despite TSMC profits souring by 40% that year, it's very hard to take anything they say seriously.
 
Daniel Nenni said:
Hi Charlie,

Apple is the one driving the process nodes into a brick wall. 20nm in 2014, 14nm in 2015, 16FFT in 2016, and 10nm in 2017. That is a very fast and very expensive pace for sure. The good news is that Apple has the money to spend on this. The bad news is that if you are not the foundry of choice you may not have the money to spend on this. Picking Samsung for 28nm, TSMC for 20nm, Samsung for 14nm, and TSMC for 16nmFFT (a different foundry for every other node) could be the strategy to keep two competing foundries in the game. If Apple sticks with one foundry for two nodes in a row the other foundry may be in CAPEX trouble. Apple certainly is the foundry king maker now.

D.A.N.
Click to expand...

I guess that another possible Apple's strategy is to inject more price and technology competition for iPhone related SoC by leveraging one foundry against the other, such as Samsung vs. TSMC. In the worst case, any issues caused by the foundry will only be tied to one generation of iPhone, about one year. Furthermore, by doing so we can always see two foundries split the Apple's iPhone SoC orders, floating between 70/30, 60/40, or 50/50. Actually this strategy may lead to a contradiction to itself. If I am a decision maker at TSMC or Samsung, what's the incentive for me to give up more profit to satisfy Apple's pricing demand? Because to make this scheme work, "both" TSMC and Samsung must guarantee to have a share. There is limited upside or downside potential to gain or lose more orders from Apple.

On the other hand, Apple might do more sourcing or even sole sourcing from TSMC for non-iPhone related, especially those projects have long lasting impacts and longer product cycles such as SoC for Apple laptops, desktop, servers, or fresh new venture such as rumored Apple Cars. One obvious reason is TSMC has little conflict of interest with Apple, compare to Samsung.
 
Last edited:
Although Apple is a driver, it is in a precarious position. Since Apple is on top, the only way it can stay there is to grow new technologies as its core technologies become common place commodities. This is why Apple is going into cars, which will be ever larger consumers of semiconductors, mems and a constantly increasing level of many technologies. Self driving, self navigating, collision avoidance, accident mitigation, entertainment, work area, energy storage, energy production, recreation, all wheel drive, and interiors designed for everything from entertainment to sleeping to temporary lodging (privacy and power generating glass) all offer a host of opportunities far greater than phones and adaptable to everything from electric bikes to any form of transport we can even imagine. This is why Apple is starting with cars, just the tip of the iceberg.

TSMC is the perfect partner with the widest variety of semi and mems technology of any company out there. They won't be the only partner, but a key one for years to come.
 
Last edited:
Goldman Sachs's view

I prefer the other quote from Goldman Sachs in the Barron's blog.

"While TSMC shares are unlikely to see strong appreciation until earnings bottom, they have tended to outperform the market at times of demand uncertainty, such as today. In addition, we estimate TSMC’s revenues will increase 6% qoq in 1Q16 as its Apple allocation continues to increase. Maintain Buy rating."

OK there may be the widely predicted slowing of growth in smartphones led by China but the approach that I find irritating beyond measure is when analysts find reasons to slate one company without looking at its competitive position. We will know a lot more in a few weeks time when the quarterly transcript is available but in the meantime:-

a. Morris Chang always said that he would lose some market share because his 16ff process was 6 months behind Samsung.
b. Apple, we are told, is working with TSM on an Apple specific version of 16ff+. I assume that Apple's cheques are not chicken feed. Will Apple give Samsung the same facility?
c. Samsung (per the rumour reported by rmaire) has cut is capex by 20%.
d. No one to my knowledge is predicting significant growth in Intel's PC market, which it looks like Tim Cook is moving into by replacing Intel in at least some of the Macs.

I am not betting against TSMC being the last man standing. Even if the tide is going out we will not see Morris Chang without his pants.
 
Just for clarification lets talk about how Apple splits manufacturing. At 28nm it was reported that there would be a 50/50 split between Samsung and TSMC. According to SemiWiki consensus it would be 100% Samsung which was the case. The same reports surfaced for 20nm and SemiWiki again predicted correctly that 100% would go to TSMC. For 14nm it has been all over the map but SemiWiki consensus is that the iPhone 6s would be Samsung 14nm LP and the iPads would be TSMC 16FF+ which is a predicted 70/30 split.

At no time has Apple split a single design between multiple fabs. The split is between the A9 and A9x SoCs.

Sound reasonable?
 
Dan,

Apple can move SoC designs between TSMC and Samsung since the processes are not that different. It would however be a costly and labor intensive process. Moving a design from TSMC to Intel would be much more difficult but again it is doable. Will Apple split a design between fabs? I seriously doubt it unless they had a compelling reason to do so. Insufficient capacity is the only reason that comes to my mind.

CharlieD
 
Daniel Nenni said:
Now that is interesting! Good catch! Could be nothing but...........
Click to expand...

As long as each iPhone model (6s and 6s Plus) has enough volume, Apple can use different A9 sub-model for each of them. Remember each phone model has different physical size and assembled at different production lines or even at different contract manufacturers. So using different A9 SoC for different iPhone 6 models is achievable.

If Apple wants to be more aggressive, they can even use different A9 SoC model for new iPhone models made for different market. Fro example A9-N for North America, A9-U for Europe, and A9-I for Asia. Apple can do anything they want!
 
Although Apple has emerged as a vital factor in the fab business, and Samsung has taken a lead in the smaller node game, TSMC's roots are much deeper in the fabless semiconductor business. Players like Samsung and Global Foundries can give TSMC a run for money, but they don't pose TSMC any existential threat. Again, because TSMC has a strong customer base and time-tested business relationships. TSMC has to catch up in the smaller node game and I am sure they will.
 
The biggest thing Samsung has going for them is nimbleness. They are incredibly nimble, which is practically inconceivable given their size. Whatever foundry market shifts occur, Samsung is likely to see them first, and respond to them first. This is due to the business intelligence they gather with all their consumer businesses. It is also likely Samsung will work to be #1 in the foundry market, as they are in many of their other businesses. They can also leverage internal demand to keep the foundry busy, as they did with the S6.

Intel's, TSMC and Samsung's business models are considerably different: Pure-play foundry; vertically integrated mfg; and chip merchant mfg. I personally think they are all respectable, and have different strengths at different points of the business cycle. I think TSMC's model has been quite effective in the growth phase of the last few years, but Samsung may have more of a cushion as we head into the downward trajectory.
 
You must log in or register to reply here.
Back
Top