Tuesday, March 30, 2010

iPhone on Verizon? Time for Some Logic

With the latest in a long line of Verizon iPhone rumors recently circulated by The Wall Street Journal, it's worth commenting on the engineering logic and how it relates to a timeline for such a potential event.

First of all, I have absolutely no insight into the Apple and Verizon negotiations or plans to the extent they even exist. Second, I don't know what Apple's contractual obligations are with AT&T, including when, if ever, AT&T's exclusivity ends. But there are at least three reasons there wouldn't be an Apple-Verizon deal of any kind in 2010 or even 2011.

1. AT&T has an exclusive until time X in the future (2012, whenever).

2. AT&T simply agrees to pay Apple more than Verizon would pay Apple in order to ensure continued exclusivity. Using round numbers, AT&T currently pays Apple $600 per iPhone and sells it for $200, eating the $400 difference upfront, taking the consumer credit/fraud risk. Verizon simply may not be willing to pay as much, or AT&T may be willing to increase its subsidy in order to extend its exclusivity.

3. Verizon may just not like Apple's terms on any one or several levels, such as those relating to the AppStore, iTunes, branding, service pricing, whatever. Verizon may be of the opinion that the iPhone would clog its networks and would require some form of variable pricing for bandwidth consumed, just like we all pay for electricity, fuel for our cars and food. Basically, if you consume more, you pay more.

But let's say that none of those three points apply. Let's say that Verizon and Apple are both free to cut a deal, and that they want to do so. What are the semiconductor requirements in order to produce this Verizon iPhone device? There are two scenarios here.

Scenario No. 1: Let's say that Apple and Verizon intend to ship the Verizon iPhone before the middle of 2011. This would mean that the device could utilize the same baseband ("radio") chip currently powering the Blackberry 9550/Storm and 9630/Tour. This is a Qualcomm chip, as Qualcomm has a de facto monopoly in the market today on chips that include the requisite CDMA/EVDO/GSM/HSPA standards for worldwide coverage. This is a product that is technically possible to ship today.

Scenario No. 2: Let's say that Apple and Verizon intend to ship the Verizon iPhone no earlier than the middle of 2011. This would mean that the device could utilize the next-generation LTE-inclusive baseband chip from Qualcomm, which was introduced at the Mobile World Congress trade show in Barcelona in February 2009. Qualcomm stated at the time that it intended to sample this chip around the middle of 2010, which normally means that it could be on store shelves one year thereafter, around the middle of 2011.

Here is a central question to what must surely be part of Apple's negotiations with AT&T and Verizon alike: Will a Verizon iPhone have the exact same hardware as the AT&T iPhone when AT&T's new iPhone version appears presumably sometime this summer? Why even ask the question?

In an ideal world for Apple, it would prefer to make only one kind of hardware, common for both AT&T and Verizon. It would mean that there would be at least a technical path to users switching between networks. It would also open up a theoretical additional path for Apple to sell a device directly to the consumer, without a service contract, just like Google does with its Android Nexus One. For consumers wanting to pay $600 or more for this flexibility, this would be a great thing.

Apple has no problem making such a universal iPhone, either under scenario No. 1 (non-LTE), or under scenario No. 2 (LTE-inclusive), or both. Same thing for Qualcomm. And yet, there are two reasons it would still be unlikely for such a common device to appear in the next couple of years.

Reason No. 1: Cost. This isn't an issue for Verizon because its device must include GSM/HSPA no matter what in order to ensure international roaming. And at that point, the cost differential between a two-band GSM/HSPA chip and a four-band GSM/HSPA chip is essentially zero. However, it's a big issue for AT&T. Why? Because AT&T doesn't want to pay the "Qualcomm Tax" resulting from Qualcomm's de facto monopoly on CDMA/EVDO. AT&T has no reason to include CDMA/EVDO in any device, iPhone or otherwise. So unless Qualcomm is willing to give away this capability for free -- which is possible, but unprecedented -- this probably wouldn't happen.

Reason No. 2: Carrier rivalry. But bluntly, it may just be the case that AT&T and Verizon both want to ensure that their iPhones each are unique to each other, even if Qualcomm was to take the chip cost off the table for AT&T. Perhaps they are both afraid of Apple's power if the same device could be used on other carriers, despite the contractual two-year subsidy lockdown.

One final point is that regardless of what AT&T or Verizon do with the iPhone for the next 12 to 15 months, Sprint, through its 57%-owned Clearwire, has a one-year time-to-market advantage with its WiMax technology. Any iPhone from AT&T or Verizon operating before the middle of 2011 simply cannot compete with the superior performance afforded by the 120 MHz worth of spectrum Clearwire has at its disposal.

The first WiMax phone for the United States is the HTC EVO 4G, introduced a week ago and available this summer. It can simply do things -- such as multiple channels of HDTV -- which no iPhone can do until the arrival of LTE in a handset as early as the middle of 2011.

And even then, keep in mind that the amount of spectrum available to both AT&T and Verizon (some 20 MHz each) pales in comparison to Clearwire's 120 MHz.

If Sprint and Clearwire can bring attractive devices to market to capitalize on this gigantic advantage, all other wireless broadband operators should be very worried. This Clearwire/Sprint WiMax network is up and running in 28 markets today (covering more than 30 million people) and will be covering at least 120 million people by the end of this year. So far, only Google and HTC (in combination) have announced that they are taking advantage of this significant time-to-market advantage afforded by Sprint/Clearwire.

Will Research In Motion's BlackBerry be the next to make the move with Sprint's WiMax advantage?