Apple may have made a huge mistake in having Samsung stamp inferior A9 chips

“One variant of the A9 is manufactured in TSMC’s 16-nanometer FinFET Plus manufacturing process while the other is built in a Samsung 14-nanometer process (a reliable source tells me that Samsung is likely using its more advanced 14-nanometer LPP process, rather than its lower-performing LPE process, here),” Ashraf Eassa writes for The Motley Fool. “Although Samsung got a lot of positive press for going into production on its 14-nanometer process before TSMC did on its roughly equivalent 16-nanometer process, I believe that Apple would have been better served relying on TSMC exclusively for the A9.”

“Both the Samsung and the TSMC processes feature minimum metal pitches of 64nm, but Samsung’s gate pitches are a little tighter at 78 nanometers versus 90 nanometers in the TSMC process. Unsurprisingly, the TSMC-built A9 chip is slightly larger than the Samsung-built A9 chip,” Eassa writes. “However, although the Samsung process is denser, the TSMC process is superior in the ways that count: electrical performance and yields … In fact, that same source informed me that TSMC’s A9 yields are twice those of Samsung’s, which would suggest that it is more cost effective for Apple to build A9 chips at TSMC than at Samsung.”

“A report recently surfaced claiming that TSMC has won the entirety of Apple’s next-generation A10 applications processor business. According to my source, this report is accurate,” Eassa writes. “Quite frankly, given that the TSMC 16-nanometer FinFET Plus process appears to be delivering better power consumption and yields, this is not at all surprising.”

Read more in the full article here.

MacDailyNews Take: The problem, and Apple’s mistake, is in how this treats the iPhone customer (and, presumably the iPad customer, too). Apple is effectively subjecting customers to a random lottery drawing. Some customers get superior iPhones and some get inferior iPhones. All at the same price via unmarked boxes. This is wrong.

Apple should have higher regard for their customers and Apple customers should expect better from Apple.

Claiming a 3% discrepancy in “real world usage” also does not address the core issue. Take time-lapse recording, a feature Apple builds into their iPhones and promotes. Here’s how an iPhone 6s performs when recording a long video – again, “real world usage” via a feature that Apple promotes – with a TSMC-stamped A9 vs. a Samsung-stamped A9:

SEE ALSO:
Apple claims iPhone 6s/Plus’ A9 battery performance only varies 2-3% between TSMC and Samsung variants in ‘real-world usage’ – October 8, 2015
Chipgate: Did you get the good A9 or the crap A9 in your iPhone 6s/Plus? – October 8, 2015

[Thanks to MacDailyNews readers too numerous to mention individually for the heads up.]

28 Comments

  1. This may be another storm in a tea cup for Apple. It looks as though both chips meet Apple’s spec, but the TSMC ones might slightly exceed it.

    It will become a big issue for some, but most customers won’t know or care which chip is inside their new iPhone.

    However it may turn into a very big problem for Samsung because it’s their chip division that is keeping them profitable, while the handset division is struggling. If Samsung are seen to be producing inferior chips to TSMC, then it doesn’t bode well for the future. Furthermore it gives Apple a very big stick to hit Samsung with as Apple can insist on paying Samsung lower prices as their products are not as good as TSMC’s and there would be a risk in continuing to use Samsung as a supplier.

    If Samsung were to lose Apple as a customer, they would be pretty well screwed. The main thing saving them is that they have more manufacturing capacity than TSMC, but there again, TSMC had minimal manufacturing capacity a few years ago, so they might soon overtake Samsung in terms of quantity as well as quality.

    Samsung and their local press have been keen to let the world know that Samsung makes chips for Apple’s iPhones, but they will now be seen as the people who make the second rate chips for iPhones.

    1. Well said, given Samsungs handset problems, i wish Apple would cut them off as chip supplier either, however what if TSMC increases production capacity but then their quality goes down?? Apple nneds both Samsung and TSMC and force the inferior chip supplier to provide better quality or get a financial penalty.

    2. People, don’t jump to any conclusion yet. Lets wait to see the facts. One video test does not prove anything. We need more sample size and test from reliable sources. For all we know, it could be one defective device. I am just surprise how quickly MDN reach this conclusion, knowing how often the media fabricate and exaggerate all things Apple. There is always variances for each chip, but if the chip really does not meet the Apple promise of 10Hrs usage, than there will be a class action lawsuit.
      Not all Intel chips perform the same either, some you can over clock higher than others, but as long as they meet a certain tolerance and specs, than its OK.

    3. Well said. Technically, this “-gate” is minutia for anyone who really understands chip fab variability, as it really comes down to what threshold criteria is being used on yield…and Samsung is not doing better here.

      In the fanboy camp, this is simply more evidence that Samsung is a hack…that’s a win for Apple, not doom.

      And from a supply chain standpoint, what’s the alternative to having two suppliers? Inadequate supply, which yields unhappy customers, a black market and another excuse for Wall Street.

    4. Since we don’t know what contracts were drawn up, it may actually already be the case that Apple is paying less for the Samsung A9s vs the A9s from TSMC. Apple could have kept Samsung for both lower cost and to keep a good supply of chips available for use.

  2. It could be a basic issue of supply constraints. If one manufacturer can only produce X chips using better technology, and you need 2X supply, you need to go to the next manufacturer for the remainder, even if it doesn’t have the same technology. As long as both, however, meet the specs for the product, there shouldn’t be an issue.

  3. Obviously most of us here would prefer TSMC to produce all of Apple’s A# chips. However it would be a huge risk having just one supplier.
    Apple may have to optimize code for the Samsung chip to improve battery life. No surprise there especially since the processor design is slightly different.

  4. MDN turn FUDmeister! Only nerds with a benchmark app will notice or care. If you need to look at numbers to see the difference, there is no difference.

    I wonder if TSMC said no, we cannot supply all the chips you want. Should Apple have stuck with Samsung for another round? Would TSMC ever get off the ground with Apple if they wait until they can do them all? Or would MDN prefer Apple launched with constrained supplies so we’d have another “botched” launch like the Watch? Maybe TSMC should have nobbled the processors to perform more in line with Samsung, I wonder at the outrage that would cause when they were found out!

    I think Apple are easing TSMC into the supply chain and the unavoidable difference we see here (when using a microscope) is of no consequence to anyone.

    Has anyone run these tests on a bunch of identical iPhones such as the 6 or two TSMC or two Samsung 6s’. See what the differences are between truly identical phones. I’d be shocked if all results came out the same, but might they show the same variance? Is there an overlap? This guy had two of each, how can he not have done this?

    And, if anyone is still reading, how does a few percent less than superb equal crap? I thought MDN considered the 6s to be superb?

    1. “Only nerds with a benchmark app will notice or care.”

      Or people who use Apple’s built-in and promoted time-lapse recording capabilities in “real world usage.” SIGNIFICANT VARIANCES IN BATTERY LIFE!!! Not a small percentage.

      People who defend Apple Inc. over even Apple customers have lost the plot completely.

      1. 30 minutes of time lapse, is that real world? And in any event you can do it, and more, you just have only 5% less battery left. Apple promoted this feature, yes, and it’s there! Fully useable on both. It is a battery draining function either way.

        No one is defending Apple over their customers. People are defending rationality, proportionality and perspective.

        Apple did not release a phone with crap performance or crap battery life. They released a phone that, in either version, everyone would be delighted with. Turns out you can measure a variance in performance from great to greater, or greater to great. The variation does not bring the phone below any claims or expectations as to performance or battery life. If no one had measured this, no one would have known.

        Let me ask this. Which came first? The observation that battery life or performance sucked, followed by an investigation to find out why and culminating in the discovery of two different processors? Or was it the discovery of the difference in processors, followed by an investigation to see if there were a difference in performance?

  5. MacDailyNews is right.

    MacDailyNews defends, in this order:
    1. Apple users
    2. Apple Inc.
    3. Apple partners (who aren’t stealing from Apple like Samsung, Google, and Microsoft)

  6. Better to have two different suppliers in two different locations than just one supplier.
    If there is a natural disaster, as happened some years ago in Japan, when a whole lot of semiconductor plant was taken down by an earthquake, Apple will not be left without a supplier.

  7. People, don’t jump to any conclusion yet. Lets wait to see the facts. One video test does not prove anything. We need more sample size and test from reliable sources. For all we know, it could be one defective device. I am just surprise how quickly MDN reach this conclusion, knowing how often the media fabricate and exaggerate all things Apple. There is always variances for each chip, but if the chip really does not meet the Apple promise of 10Hrs usage, than there will be a class action lawsuit.
    Not all Intel chips perform the same either, some you can over clock higher than others, but as long as they meet a certain tolerance and specs, than its OK.

Reader Feedback

This site uses Akismet to reduce spam. Learn how your comment data is processed.