Apple A9 chip fab TSMC reports record earnings, casting further doubt on 'Peak iPhone'
Taiwan Semiconductor Manufacturing Co., the world's largest chip fab and a manufacturer of Apple's A9 chips used in iPhone 6s models, has posted earnings that beat analysts estimates for the quarter and established the company's highest annual earnings in 29 years.
TSMC
According to a report by Focus Taiwan, TSMC reported NT$306.57 billion ($9.15 billion US) in net profit for 2015, a 16.2 percent annual increase and a record high for the firm. Its consolidated sales rose 10.6 percent to NT$843.497 billion ($25.2 billion US), also a new record. About $3.7 billion of TSMC's revenue comes from Apple.
For the December quarter, TSMC exceeded analysts' average estimates of NT$68.5 billion to reach NT$72.8 billion ($2.2 billion US) in net income, according to a report by Tim Culpan of Bloomberg.
Culpan stated that TSMC's record quarter comes alongside "a move to more sophisticated manufacturing technology," and noted that Apple had also forecast record revenue, "fueled by customer upgrades and sales in China, even with the Asian country's economic slowdown."
Bloomberg also referenced data from the Chinese government noting a 33 percent increase in non-Android smartphone sales, "the vast majority of which would be iPhones," which AppleInsider highlighted on Tuesday.
It also cited Ethan Chen, an analyst with Sinopac Financial Holdings, as saying "We think TSMC's wafer shipment will regain traction on a quarterly basis in 2Q16 thanks to regaining most allocation for Apple A10 chip orders."
In September, TSMC was rumored to have been selected as Apple's exclusive manufacturer for this year's A10 chip expected to power iPhone 7 and new iPad models.
The new chip is expected to use a 16-nanometer process combined with a new InFO (integrated fan out) architecture, which allows chips to be stacked on top of each other and mounted directly to a circuit board, instead of onto a substrate first, reducing both the thickness and the weight of devices. Apple is rumored to be TSMC's first customer to use InFo, and A10 production is expected to begin in March.
Samsung had been Apple's sole source for ARM-based processors powering iPods and iOS devices going back to 2006, the year Apple dropped PortalPlayer and moved to a Samsung chip for the fifth generation iPod.
Apple subsequently used Samsung-designed chips in early iPhones (after Intel expressed no interest). It embarked upon an internal chip design effort in 2008 that included the acquisition of fabless chip designers PA Semi and Intrinsity.
Source: TechInsights
That effort culminated in the release of A4, Apple's first custom designed ARM processor, which was used in 2010 to power the new iPad as well as iPhone 4 and a new iOS-powered Apple TV. The company continued to rapidly advance upon its custom chip designs, resulting in the 2013 release of its pioneering 64-bit A7, a new architecture that sent Qualcomm scrambling to accelerate its own chip roadmap.
Apple has recruited chip design talent away from Texas Instruments (which exited the consumer market in 2012 after a string of Android product failures doomed its OMAP chips), along with other struggling chipmakers including AMD, IBM and Freescale.
As Apple's chip appetite grew--and as its chip partner Samsung increasingly promoted its own Exynos processors--the company courted TSMC as a new supplier for the A8 chips that would power iPhone 6.
TSMC had more advanced chip fabrication technology than its South Korean rival up until 2009, when Samsung poached disgruntled TSMC veteran Liang Mong-song to lead its own System LSI chip fab as a Chief Technology Officer. After illegally delivering TSMC trade secrets to Samsung, Liang's System LSI launched a rival 14nm FinFET process superior in some respects to TSMC's.
While Apple shifted most of its 20nm A8 production to TSMC, it used Samsung as a primary supplier for about 60 to 70 percent of its A9 chips. Samsung's 14nm process resulted in smaller A9 chips than the 16nm FinFET process used by TSMC, although the latter appeared to perform better. Tests showed that in real world use, the actual difference between the two batches of A9 chips was marginal.
This year, Apple is expected to exclusively use TSMC and its new InFo architecture to build the A10. In parallel, Samsung has taken over Qualcomm's future chip fabrication from TSMC, the only foundry that has ever produced Qualcomm's high-end Snapdragon chips.
However, with Apple dominating premium, high-end smartphones and tablets and with sales of higher end Androids shrinking and fragmented between Qualcomm, Samsung's own Exynos, and other in-house chip designs from Huawei and LG, Qualcomm appears to be at risk of suffering the same inability to risk investment in expensive new chip technology that earlier forced Texas Instruments and Nvidia out of the smartphone chip business.
Source: Antutu
Antutu benchmarks already show that Apple's A9 is significantly faster than the current chips from Qualcomm, Huawei and Samsung, even when those chips are clocked higher and paired with more RAM (albeit also running Google's less efficient Android software).
TSMC
According to a report by Focus Taiwan, TSMC reported NT$306.57 billion ($9.15 billion US) in net profit for 2015, a 16.2 percent annual increase and a record high for the firm. Its consolidated sales rose 10.6 percent to NT$843.497 billion ($25.2 billion US), also a new record. About $3.7 billion of TSMC's revenue comes from Apple.
For the December quarter, TSMC exceeded analysts' average estimates of NT$68.5 billion to reach NT$72.8 billion ($2.2 billion US) in net income, according to a report by Tim Culpan of Bloomberg.
Culpan stated that TSMC's record quarter comes alongside "a move to more sophisticated manufacturing technology," and noted that Apple had also forecast record revenue, "fueled by customer upgrades and sales in China, even with the Asian country's economic slowdown."
Bloomberg also referenced data from the Chinese government noting a 33 percent increase in non-Android smartphone sales, "the vast majority of which would be iPhones," which AppleInsider highlighted on Tuesday.
Exclusive on Apple's A10
It also cited Ethan Chen, an analyst with Sinopac Financial Holdings, as saying "We think TSMC's wafer shipment will regain traction on a quarterly basis in 2Q16 thanks to regaining most allocation for Apple A10 chip orders."
In September, TSMC was rumored to have been selected as Apple's exclusive manufacturer for this year's A10 chip expected to power iPhone 7 and new iPad models.
The new chip is expected to use a 16-nanometer process combined with a new InFO (integrated fan out) architecture, which allows chips to be stacked on top of each other and mounted directly to a circuit board, instead of onto a substrate first, reducing both the thickness and the weight of devices. Apple is rumored to be TSMC's first customer to use InFo, and A10 production is expected to begin in March.
Apple's shift from Samsung to TSMC
Samsung had been Apple's sole source for ARM-based processors powering iPods and iOS devices going back to 2006, the year Apple dropped PortalPlayer and moved to a Samsung chip for the fifth generation iPod.
Apple subsequently used Samsung-designed chips in early iPhones (after Intel expressed no interest). It embarked upon an internal chip design effort in 2008 that included the acquisition of fabless chip designers PA Semi and Intrinsity.
Source: TechInsights
That effort culminated in the release of A4, Apple's first custom designed ARM processor, which was used in 2010 to power the new iPad as well as iPhone 4 and a new iOS-powered Apple TV. The company continued to rapidly advance upon its custom chip designs, resulting in the 2013 release of its pioneering 64-bit A7, a new architecture that sent Qualcomm scrambling to accelerate its own chip roadmap.
Apple has recruited chip design talent away from Texas Instruments (which exited the consumer market in 2012 after a string of Android product failures doomed its OMAP chips), along with other struggling chipmakers including AMD, IBM and Freescale.
As Apple's chip appetite grew--and as its chip partner Samsung increasingly promoted its own Exynos processors--the company courted TSMC as a new supplier for the A8 chips that would power iPhone 6.
TSMC had more advanced chip fabrication technology than its South Korean rival up until 2009, when Samsung poached disgruntled TSMC veteran Liang Mong-song to lead its own System LSI chip fab as a Chief Technology Officer. After illegally delivering TSMC trade secrets to Samsung, Liang's System LSI launched a rival 14nm FinFET process superior in some respects to TSMC's.
While Apple shifted most of its 20nm A8 production to TSMC, it used Samsung as a primary supplier for about 60 to 70 percent of its A9 chips. Samsung's 14nm process resulted in smaller A9 chips than the 16nm FinFET process used by TSMC, although the latter appeared to perform better. Tests showed that in real world use, the actual difference between the two batches of A9 chips was marginal.
Qualcomm jumps from TSMC to Samsung
This year, Apple is expected to exclusively use TSMC and its new InFo architecture to build the A10. In parallel, Samsung has taken over Qualcomm's future chip fabrication from TSMC, the only foundry that has ever produced Qualcomm's high-end Snapdragon chips.
However, with Apple dominating premium, high-end smartphones and tablets and with sales of higher end Androids shrinking and fragmented between Qualcomm, Samsung's own Exynos, and other in-house chip designs from Huawei and LG, Qualcomm appears to be at risk of suffering the same inability to risk investment in expensive new chip technology that earlier forced Texas Instruments and Nvidia out of the smartphone chip business.
Source: Antutu
Antutu benchmarks already show that Apple's A9 is significantly faster than the current chips from Qualcomm, Huawei and Samsung, even when those chips are clocked higher and paired with more RAM (albeit also running Google's less efficient Android software).
Comments
Yeah, ripping off the iPhone patents really paid off for Samsung, didn't it.
Never bite the hand that feeds you. Even my cat knows that, and I didn't even have to sue him.
Be skeptical if something becomes "common knowledge" in the media...
Anyway JamesBB, i notice that after three months of selling, last week the 6s had 9.8% penetration (cf 6 @ 31.4%) and 6s plus is 3.9% (cf 6 plus @ 11.3%). Interesting. I would say the 6s's are selling just as well, if not better, as they have only been for sale about 20 per cent of the time of the 6 and 6 plus. I think eventually though they will probably sell less, as you say, while overall iphone sales will be higher.
Heck the senior citizen entropys just bought two 6, to my 6s and mrs entropy is sticking with her 6 plus for the time being.
As for "peak iPhone" well that is bound to happen sometime. Every product ever conceived by man eventually finds market equilibrium. I serious doubt though that this will happen in 2016.
On a side note I'm close to becoming a one percenter. Still rocking an iPhone 4. I'm not sure if it will make it to the next iPhone release though. IPhone 7 sounds very interesting though it is likely to be too big for my tastes.
And then you have the loss of Qualcomm chips to Samsung on top of that...!
I am actually surprised they are in such good shape considering all the horrible news (where none is related to Apple...).
Bloomberg (and CBNC, WSJ, BI etc) are all prostitutes of big guns on Wall Street, so obviously they would assign all of TSMC's problems to Apple without a droplet of factual data... and the others representing 85% of their business are doing just fine in the traditional PC industry!
Come one... listen to yourselves...
The same thought process can be applied to peak iPhone. To make any kind of prediction one must assume Apple will continue to build out distribution to cover all the geographies and markets currently covered by all smartphone vendors. Samsung, for example, has some 900 carrier partners around the world, to Apple's approximately 350. As Apple adds carrier partners, it introduces the iPhone into markets where Samsung and other phone vendors haven't yet had to compete on a level playing field against Apple. Apple's entry simultaneously adds a new market for Apple while introducing a strong competitor to everyone else, at least at the profitable premium end of the market. One must also assume Apple might do as it did in the iPod line, and it's other product lines, extending the iPhone line to cover a wider range of prices. Today, Apple sells only flagship iPhone models. This year's and the flagships of the previous couple years. It has not yet experimented with extending its iPhone line to include a mid-priced iPhone. But it could, and that would push out the peak iPhone curve significantly.
Phones, tablets, watches (or gadgets in whatever form factor) are consumables. You can argue whether phones are discretionary anymore (I certainly don't consider my phone discretionary). They are continuously replaced in fairly steady cycles although they may evolve and change form factor.
Markets will grow VERY fast as long as you have relatively unexplored markets (which we still do - think Africa, India, Asia etc).
When markets mature, then you enter replacement cycles where growth starts to taper off. Growth in this phase can still be high (think car industry during the past decades), but then it's usually driven by improvements in technology, in the use of products or demographics (like the growth of the Chinese middle class).
And we haven't even touched on market share... with just ~15%, guess Apple still has plenty of room to grow on that account. For every new Android feature phone, there is a potential iPhone customer with a steady 94% replacement loyalty.
What "peak" are we referring to here?