For the past year I have been thinking that in a rational world Apple (AAPL) and Intel (INTC) would have established a manufacturing partnership as soon as Intel's 22nm Trigate manufacturing process was announced in 2011. Yet they haven't. Speculation and rumor run rampant. Articles on Seeking Alpha and elsewhere offer perfectly reasonable arguments, for and against, Intel manufacturing for Apple. One thing that everyone agrees is that the Apple/Samsung foundry relationship is not sustainable.
Yesterday I was poking around on the Internet and it came to me! Memory. It's all about memory.
Let me explain:
Every Application Processor in every smartphone is part of a PoP (Package on Package) assembly. The PoP assembly is made up of the Application Processor on the bottom and two Mobile DRAM chips mounted on top of the processor. The reason for this is often thought to be for the required miniaturization, which is true, but this kind of assembly contributes a great deal to the speed performance and the low power requirements of mobile devices. Tablet computers, where real estate for mounting components, is much larger still use PoP assemblies for the performance reasons cited above. Smartphones and tablets can't be made without mobile DRAM.
The mobile DRAM business is made up of Samsung, Hynix, Elpida, and very little from Micron (MU). The consensus a year ago was that Elpida would go out of business, leaving Samsung and Hynix as an effective duopoly in mobile DRAM business.
To Tim Cook, CEO of Apple, a duopoly made up of arch enemy, Samsung and another Korean company, in a commodity that will require 600 million devices to be shipped to Apple is simply unacceptable.
Fast forward to today; Micron is in the final stages of acquiring Elpida. Check the timing on the following events: on February 27, 2012, Intel and Micron announce an "enhanced" joint venture that adds other memory technologies. Part of the enhanced joint venture seemed to beef up Micron's balance sheet to prepare for the Elpida acquisition. On May 8, 2012, it was announced that Hynix, Toshiba and others had walked away from the deal and Micron had won the competition to buy Elpida out of bankruptcy. On May 16, 2012, rumors emerged of large Apple mobile DRAM orders being placed with Elpida at the expense of Samsung.
Without a reliable and capable supply of mobile DRAM there is no way that Apple could move the production of "A" chips away from Samsung. At least for the time being Apple could depend on Samsung to continue supplying, since Apple is thought to have largely funded the Samsung factory in Austin, Texas.
The Bondholders of Elpida have about two more weeks to file an appeal to the Japanese bankruptcy court approval of the Micron acquisition. After that time expires, Micron believes they can complete all the paperwork in another 30 days and close the Elpida acquisition.
So, about May 1, 2013 there should be noisy celebrations at Apple, Intel, and, of course, Micron.
After the close of the Elpida acquisition, Apple will be free to move the "A" chip fabrication anywhere they desire. They will desire to put the business with Intel because any other potential supplier is a technology dead end. Intel will have access to a long trusted partner, Micron, to cooperate in supply and development of new memory technology for mobile, server, and storage application.
I expect the second half of 2013 will be very interesting for Apple, Intel and Micron.
Buying any or all of these stocks should be very rewarding in the next 18-24 months.
Disclosure: I am long INTC, MU. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Source: http://seekingalpha.com/article/1280991-why-intel-isn-t-building-the-apple-a-chips-yet?source=feed
new world trade center kellen moore octomom stoudemire jordan hill tony nominations dark knight trailer
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.