AT&T appears to have had a change of heart about HSPA+. Last year, it said it would shelve the technology to focus on its upcoming long-term evolution deployment in 2011, but in an interview with GigaOM, AT&T CEO of Operations John Stankey said the operator would upgrade its 3G HSPA networks to support theoretical speeds of 14 Mb/s and real-world speeds of 7 Mb/s. There’s some confusion over whether this is really HSPA+, which has a theoretical ceiling of 21 Mb/s, but either way AT&T seems set on at least doubling its capacity from its current 7.2 Mb/s network, regardless of whether its technically doing HSPA+.
Why the reversal? According to Stankey, the upgrade is cheap — less than $10 million to cover 250 million pops — and the LTE ecosystem that will deliver a wide range of dual- or tri-mode handsets won’t really be in place until 2014. Verizon Wireless, however, seems to be refuting that claim. Lowell McAdam (who has been talking a lot about LTE lately) said at the Reuters Global Technology Summit in New York that VZW would have up to five handsets available on its 4G LTE network by this time next year. He wasn’t just tossing out wild estimates either. He named specific vendors: Motorola, HTC, LG Electronics and Research In Motion are developing handsets. Samsung has said it would unveil its first dual-mode CDMA 1X-LTE handset this year for MetroPCS.
So who’s right? Will LTE be a handset wasteland for several years, offering up only USB dongles and the occasional embedded laptop, or will it be a handset- and tablet-driven frenzy?
Connected Planet’s take,
Kevin Fitchard:
It’s not just AT&T that’s changing its tune here. You may recall that VZW wasn’t that high on the 4G handset for some time. While it said it would definitely offer handsets over the new LTE network, its biggest hopes lay in a new hyper-connected world where everything from your Internet tablet to your Toyota Prius have an LTE chip embedded. With that concept in mind, Verizon said it would focus on broadband connectivity to laptops, netbooks and other bandwidth hungry devices initially, with an eye on the bigger picture down the road. Meanwhile the 3G EV-DO network would carry the smartphone load. That strategy remains unchanged, except for VZW placing greater emphasis on the 4G smartphone. It could just be McAdam is throwing a bone to the market and the financial press — smartphones are the hot thing right now. Even if handsets aren’t the bedrock of VZW’s 4G strategy, it doesn’t hurt to score a few points by touting them now.
If Verizon — and even tiny MetroPCS — has commitments for LTE-CDMA handsets a year before AT&T’s own LTE network goes live, AT&T should have no trouble procuring commitments from vendors of its own. Stankey is right in the sense that AT&T will be among the first GSM operators to deploy LTE and will have to wait for the rest of the industry to catch up. But like Verizon, AT&T is large enough to drive the market on its own. Maybe its Android-powered 4G-3G handset will be more expensive than the one Vodafone offers two years down the road, but if it wants it dozens of device-makers will trip over themselves to provide it. And when we start talking about data-only 4G devices, AT&T can hitch a ride on VZW’s supply chain. Both their LTE networks will use the same frequencies, meaning that for the first time they’ll have interoperable networks.
In my opinion, both AT&T and Verizon are adjusting their 3G and 4G strategies to account for the new reality of the smartphone. A year ago, it probably made sense for Verizon to rely on 3G to power its handset and smartphone services while tapping 4G for workhorse broadband connections and new business models. A year ago, it probably made sense for AT&T to think that 7.2 Mb/s HSPA was plenty to handle today’s smartphone data demand and focus its energies and capex on the new LTE network. But smartphones have taken on a hunger of their own. While they consume only a fraction of the data as a dedicated laptop card, there are also millions more of them on both operators’ networks — that adds up. Meanwhile, more-sophisticated OSes and applications are doing much more than just Web browsing, increasing the demands that each individual device places on the network.
AT&T and Verizon are using different networks to solve what essentially the same problem. With LTE around the corner, VZW has no interest in an upgrade to an advanced 3G technology like AT&T, so it plans to shift a good portion of growing smartphone traffic to the 4G network. Meanwhile, AT&T still has well over a year to wait before it can take the same step, so it’s bulking up its 3G network to handle the anticipated traffic explosion. As Stankey pointed out in the GigaOM interview, the HSPA+ device ecosystem is almost as underdeveloped as the LTE one. But AT&T doesn’t necessarily need faster phones, it needs a higher capacity network. Sure, each iPhone might be limited to the 7.2 Mb/s its chip could theoretically support, but AT&T could support twice as many of them in each cell.
One last thought on AT&T’s upgrade: While the term HSPA+ is being tossed around, I suspect Stankey is really talking about the final iteration of HSPA, not the new and improved HSPA+. Plain vanilla HSPA can be upgraded to 14.4 Mb/s, which is in line with Stankey’s comments to GigaOM, and it would presumably be much cheaper than a full-bore HSPA+ upgrade. When AT&T first deployed its 3G networks mid-decade the HSPA+ standard hadn’t yet been passed. So while its networks should have been hard-wired to support all of the future HSPA upgrades, not so for HSPA+. For the price of $10 million, Stankey is talking about a relatively simple software upgrade, but to do HSPA+ AT&T needs to mess with the baseband modulation of all of its base stations, something its equipment wasn’t originally designed for, likely making it a much more capital-intensive endeavor.
That’s our take on this. Let us know what you think in the comments section below: