Wireless Industry Analyst

You know, the one thing that annoys me with it being based on the SGOLD-3H based is that it has been sampling since late Q1/early Q2 2006 and I think mass production started in Q4 2006. Baseband certification and integration takes a long time, but Apple is also well known for moving faster than their competitors.

That doesn't make it impossible at all, but it'd certainly make Steve Jobs' claims about battery life nothing more than a lie. Alternatively, maybe it's a newer chip in the SGOLD3 Family, although my impression (but I'm not sure) is this one is already on 65nm...

I don't think Apple is prone to much vendor loyalty, but I still very clearly agree with you that it is most likely an Infineon chip, and I'd also tend to think it likely is the SGOLD-3H. I just find it strange though; I know this kind of thing can take a long time, but that's a bit extreme...

About one year ago, I noticed that the iPhone 3G timeframe would match perfectly for Icera's 65nm chip, assuming qualification could be done more rapidly than usual. I asked someone with some experience in qualification and apparently the timeframes I gave him seemed reasonable. However, Icera hasn't given any news whatsoever regarding the status of their chip for 6+ months...

I also noticed a Icera-Sirific news piece saying this: "The American phone company AT&T has already signed up for Icera-based cards, while Orange is also understood to have certified Icera's kit." - this is probably just coincidental, but one of the most obvious links between AT&T and Orange is they're both exclusive iPhone carriers. And the fact there is no reason given for why it's certified at Orange (unlike AT&T there's no data card deal?) makes me ponder.

I'm probably just reading too much into this. Either way I don't think Qualcomm or Broadcom are very likely for the iPhone 3G; My opinion/guestimation is that there's a 80% probability for Infineon and a 20% probability for Icera. Obviously such a deal would be even more important for the latter than for the former - but also unlike with InterDigital, there's no direct way to invest based on it since it's pre-IPO.

BTW I'm curious - why do you think they stuck with Marvell for WiFi? I guess the only disruptive technology worth switching to is Atheros', maybe you're just trying not to be too biased towards your own company... :) The AR6002 started mass production in Q1 according to PRs, that's a bit late but might be just enough. Of course, that kind of thing is just guesswork. Same for the app processor.

Oh, and sorry for the long comment, looks like I got carried away again... ;)

Arun,

No worries on the length of the comment. I truly appreciate the insights you bring in with your comments. I have always wanted my blog to be a forum where people can bounce ideas off each other about the wireless value chain which is a complex beast.

Going back to your SGOLD-3H points, I think that getting the baseband certification and integration will be a easier than starting from scratch. The reason is that the MP-EH platform is very similar to the MP-EU platform. So integration should be better than when you start from scratch, especially if you change the platform. Note that I have taken a platform centric view rather than a baseband chip-centric view.

I realize (and know from what I see in the WiFi space) that Apple is not vendor loyal as much as it worries about performance and technology leadership. But the platform view of things and the effort that has gone into characterizing the first iPhone will not go wasted, not in the time-frame that Apple was looking to launch 3G following its predecessor.

It is also for this same reason that I think Marvell will retain its WiFi slot. Again, Broadcom can displace it with a bundle and that is really best shot anybody can have at Marvell. But this being said, remember that the volumes and the margins we are talking about currently will make an incremental revenue impact on the WLAN provider.

As for Atheros, I do think we have a better mobile WLAN solution than Marvell and perhaps the best in the market in terms of the performance/power curve. But for the same reasons stated, Apple will likely not consider us for this version. (Let me take this chance to mention that I work for our networking (PC/laptops)business and deliberately try to keep away from my company's business details on the mobile side to avoid a conflict of interest. So these are only my opinions as an analyst following the developments, not as an Atheros employee)

With respect to Icera in iPhone, I am not sure if you chanced on a three-part series I wrote on the iPhone 3G suitors back in June last year. I discussed Icera at length from an iPhone perspective. I too see a lot of synergies here, but unfortunately the lack of a complete platform, and of course the lack of proven success as well in the mobile phone space, makes its solution a tougher sell.

I do think that Apple will open its doors for other players in the next couple of years, moving itself more towards performance and feature integration. That is when a proven innovator like Qualcomm or an up-and-coming Icera can win the battle.

With on2 Technologies codecs in the Infineon chips, i would think they are going to be a big player in the video codec space, not to mention that ON2/Hantro chips are also being integrated into Nokia Samsung etc. Meditek that is expected to get a major chunk of the business from China ,is also a major customer of ON2 technologies. Skype. Youtube, Myspace, Monsoon Multimedia. Move etc , to name a few have also incorporated the ON2 codecs. The talk of the town for video architecture should be ON2 Technologies.