3G iPhone using same 90nm Application Processor as 2G iPhone?

Friday 13th June 2008, 02:48:00 PM, written by Arun

We published an article analysing the likely components of the 3G iPhone, and we stand by just about everything we said now that Apple made the announcement public; however, there is one big exception: the application processor. It looks like it's actually the exact same as in the 2G iPhone. Hmmm!

The evidence is simple: the probability that battery life for both video and audio is the exact same (7 & 24 hours respectively) despite using a different application processor on a newer process node (and likely the same battery size) is basically zero. However, as witnessed by our wireless components analysis, Apple's lead times are still class-leading and 65nm would have been a perfectly realistic process node in this timeframe. Therefore, the most likely conclusion is simple: Apple's engineering operations simply didn't have the scale to deliver a new application processor so soon after the first one.

This gives further credence to the claim that PA Semi was acquired not for their roadmap or even their CPU expertise, but only for their engineering talent. Grabbing hundreds of talented engineers in one go when it seems your end-product roadmap is being limited by your semiconductor team's size looks like a very reasonable thing to do. Alternatively, perhaps they just failed to execute and had to decide to stick to their old application processor shortly before launch, but we do not believe that scenario to be very likely.

This brings up another question however: what is Apple's application processor roadmap, and how does it tie in with their overall product roadmap? It seems rather unlikely for the 3G iPhone to be replaced in the short-term, although if the only change was better batter life and initially unexposed better 3D Graphics, it doesn't seem completely unrealistic as a 1Q09 refresh on 65nm.

Alternatively, their next application processor may be initially (or even primarily) targeted at other product lines such as the iPod Touch or an iPhone Nano. The only major changes in the 3G iPhone seem to be 3G, GPS, and better Bluetooth. So this brings up the question of what, if anything, could make a new iPod Touch better than last year's version; GPS wouldn't really cut it, especially given that it couldn't be assisted by the cellular network and so wouldn't be very useful indoors. So perhaps the new application processor will be first released in that product line, and only introduced in the iPhone post-holidays?

Apple often works in mysterious ways, and trying to predict their roadmap in precise terms with little factual evidence is doomed to fail. In conclusion, it is clear that this development is both a surprise to us and quite a few other people who had predicted Apple would synchronize the 3G iPhone's announcement with a 65nm application processor refresh. The possible reasons are quite obvious, but the consequences on their roadmap are much less clear and we look forward to new data points in the future that might shed some light on this question.


Discuss on the forums

Tagging

apple ± iphone, 3g, 90nm, 65nm, 45nm

Related apple News

iPhone 3.5G to support 1080p H.264 High-Profile Dual-Stream Decoding?
Apple acquires PA Semi; Press proves lack of understanding. [UPDATED]
Apple picks AMD Radeon HD GPUs for iMac
EA games for Mac announced at WWDC not native
NVIDIA and Intel power the Apple TV