Just hours before Apple is set to release its latest quarterly earnings, it decides to steal the march, and announces the acquisition of P.A. Semi, a fabless semiconductor house that focuses on the PowerPC architecture. P.A. Semi is a fairly small company, about 150 people, and puts its main efforts into the low power desktop PowerPC market. The deal supposedly closed at $278m.
So, why would Apple decide to buy a design house? Some of the theories are:
1. Reduced cost for IC for its iPod & iPhone products by adopting in-house developed IC. This makes less sense since P.A. Semi focuses on products such as dual core, 2GHz PowerPC chips with 5-13w power consumption. In other words, it would be pulling out your cannon to kill a bird.
2. Acquiring a design house will help Apple in the software market, due to better software products. This is probably true, however, it is dangerous to acquire a design house only to tap into the staff's IC design knowledge to make better software. After all, the engineers could just walk out the door and go elsewhere or start their own firm again.
3. Ditching Intel and the x86 platform. To me this sounds too unlikely. If Apple decided to switch back to PowerPC, it could create some nasty challenges with ISVs that have been working on their x86 transitions. If Apple could ditch x86, and get away with it, I am sure they would. Then Apple could do more benchmarketing against Windows PCs. Then again, with all the issues around Vista and the lack of interest in the operating system, does Apple need more ammunition?
But, what else could there be?
1. Apple wants to breath life back into Power, and does so by acquiring P.A. Semi. Expect Apple PCs to not only be equipped with Intel x86, but also with PowerPC once again. I would assume there will be more details about this in a few hours, but also at the World Developer Conference in June this year. Apple has a tight grip on its supply chain, and strengths its grip by designing its own PowerPC chips. This move would fit into computers aimed at markets with lower computing power needs, such as MacBook Air.
2. Following 1, a next step would be to switch Apple TV to the same Power architecture, while at the same time they add new features that make it more compute intensive.
3. Increase the power effiency and adopt the architecture for iPhones/iPods/other embedded devices. This could in the long term be successful, but sounds hard to accomplish given the entrenchment of ARM processors in this market.
The next few months will prove more insights as to where this is going, and it now looks more likely that Apple will use two processor designs for its compute intensive products going forwards.
