Back in Mar 2012 Apple brought out the iPad with its A5X dual core quad graphics processor in it Fabbed by Samsung still with their old 45 nm non HKMG gate oxide process. The result: a larger die, more heat dissipation requiring a heat spreader and clumsier spread out arrangement with slower SoC to Memory access.
Just 2.5 months later in Jun 2012 Samsung brought out their own Galaxy S3 with a Quadcore SoC built by their latest 32 nm HKMG gate oxide. Result : smaller and faster die uses only 1200 mWatt, no heat spreader, memory stacked over SoC, higher bandwidth.
Could Samsung not have hurried up their 32 nm process to build the A5X for Apple and save Apple a lot of technical hassles ? Did they deliberately hold it back from Apple, their largest Customer & Competitor ?
What does Wall St. have to say about Apple continuing to depend on Samsung, their largest competitor for a key component like the SoC which has such major impact on system performance ?
First, don't see evidence that Samsung deliberately hold it back from Apple; second, even they did, there is nothing wrong in it. Apple must have been well aware of this when they made the decision unless you want to say Apple is stupid
Samsung and Apple are too big as suppliers/customers not to work with each other.
But given the fierceness of their handset competition, Apple would be wise to find alternative key component suppliers if it can find any who can deliver what Samsung can.
Appple could have found alternative key component suppliers instead of Samsung but so far They have not. What this tells you is Samsung has probably the lowest cost of production for Apple and the moment Apple switches to other suppliers, either your iphone price will go up or Apple won't make as high margin as it used to. I bet it is a big headache for Apple right now.
The patent swamps are filled with alligators, patent lawyers, pharma companies, a sprinkling of engineering, and elsewhere it is mined, so careful where you stand or step. What was good in the late 1700s no longer serves us now. Pharma companies want protection for as long as possible, but business patents, software patents, and many electronics patents outlive their technology. Even though prior art exists in many electronics patents, the legal fees are too high to risk court cases for many, who simply settle. It should be easier to invalidate poor patents, but again, the cost is too high, which is not something engineers can solve, and the legal guardians have no incentive to. We will visit this same mess in ten years time.
A nice balanced view of both sides. I think at the end of the day, it seems that nowadays that the gladiators of old are replaced now by armies of lawyers doing pretty much the same thing - bloodying the competition into submission or death irrespective of any fundamental moral rights or wrongs that may exist. Reams of arrows before are replaces by reams of money now, and the one who has the most, usually wins.