Embedding Wi-Fi in laptops created a seachange in communications, by giving most people their first experience of something approaching ‘any time anywhere’ broadband, and hugely raising expectations of how much bang for the buck a wireless data network should deliver. This blew most of the early business projections for 3G out of the water, in terms of acceptable data rates and tariffs, but while the cellcos struggled to adapt to that unwelcome fact, and users embraced Wi-Fi hotspots and hotzones eagerly, the Wi-Fi operators themselves were even further away from a profitable business model.
Now, painfully, the 3G providers are accepting the reality of the flat rate, ubiquitous broadband world and, with more attractive pricing and usability and wider availability of 3G and even 3G+, they are fighting back against the hotspot.
Another development was needed to ensure that Wi-Fi would be a stepping stone on the way to mobile broadband, rather than a long term solution, and that was 3G in laptops ‘ at least until smartphones become truly usable as computers. This is on the rise, and this year will see embedded 3G becoming commonplace in new notebook models, removing one of the last advantages that Wi-Fi can offer to the user with a 3G data account (and those are the users all operators need to chase, rather than casual hangers-out in coffee shops).
Both vendors and operators still need to adjust to these trends though. With Wi-Fi largely the vehicle for Intel’s bid to dominate the mobile broadband market, the chip giant will have key decisions to make if it sees 3G gaining an upper hand ‘ notably whether to keep its hopes fixed entirely on WiMAX as the next iteration, or use its strengths in the notebook sector to try to carve a place for itself alongside Qualcomm and TI. And to take mobile broadband users fully back under their wings, from the no-man’s land of Wi-Fi, cellcos will need to make concessions in their laptop data and roaming policies just as they are doing in voice and data on the phone, in order to present the most attractive experience and pricing to users.
Embedded 3G has been a key step on the way to making cellular networks a genuinely viable option for the all-important business traveller market, the one on which Wi-Fi laptops made their fortune (even if the operators found it hard to make profit on a data-only model).
Unlike them, cellcos, however much pressure their traditional businesses may be under, do have the advantage of brand awareness, still-lucrative voice revenues and multiple service bundling options to cushion the blow of having to grasp the nettle of flat rate pricing and reduced roaming fees. However, their next move will have to be to make roaming even cheaper and easier, and to introduce hotspot-style casual use mechanisms, so that 3G laptop owners are not restricted in their choice of network, or faced with exorbitant charges every time they log onto another carrier’s system.
There is a long way to go, but operators like T-Mobile in Germany and Carphone Warehouse in the UK have already experimented with subsidizing laptops in the same way they do phones, and AT&T is one of the cellcos talking up plans for prepaid data plans and SIM cards, and hourly access rates, for non-customers in future.
Other obstacles in the way of a truly universal mobile data experience have included the CDMA-UMTS division, and the need for laptop makers to become like phonemakers, and undergo multiple carrier certifications to include 3G in their products.
Clearly the latter does not fit well with the PC model, with its highly commoditized approach and thin margins, and of course, even in the device arena there are the beginnings of a move towards open devices that can be used on any network with minimal effort from the manufacturer or user.
This may, in reality, be far away for phones, but it would certainly be seen as a backward step to introduce complex certifications for laptops. Both these lingering challenges have so far made the external PC card or USB dongle a more viable option than embedded 3G, even though the Wi-Fi experience makes it very clear that, once users have embedded capabilities as a standard feature of their new PCs, they are inclined to use these, even if they had not seen the need before ‘ so usage soars. This, of course, is Intel’s thinking, and that of WiMAX operators like Sprint, in looking to embed WiMAX in notebooks and other devices with heavy data usage at an early stage.
The two problems with embedded 3G are starting to be addressed, most notably by Qualcomm with last year’s announcement of its Gobi product, its most aggressive assault to date on Intel’s laptop heartland and on the dominance of Wi-Fi (and potentially WiMAX) in business mobile data.
Qualcomm may have overcome its initial hostility to Wi-Fi to incorporate the technology very effectively into its integrated multi-radio handset chipsets, but the platform remains anathema to the chip giant’s broader views on technology and intellectual property (and it is no doubt hoping that a strong showing for embedded 3G will make WiMAX redundant in notebooks, and avoid any need to bow to the market and support 802.16e in future, even though this remains officially an option for Gobi).
Gobi is a combination of firmware and chips (primarily a software defined radio and a GPS chip), and is initially targeted at notebooks, but potentially suited to other devices in future too. It represents a limited but important step on the road towards software defined, fully adaptive radio products that can move between different networks and frequencies as required, transparently to the user, and without huge additional cost to the device maker.
Currently, it supports W-CDMA/HSPA and CDMA EV-DO Rev A, already promising to overcome roaming problems between the two main 3G networks, and produce improved economies of scale, particularly for the CDMA manufacturers, which are dealing with the risk of a possible decline in their market. In future, Qualcomm says Gobi will be able to support other radio technologies too, including HSPA+, LTE and future versions of the CDMA platform, plus even WiMAX if the market demands it, or other technologies like digital TV transmitters.
Gobi can be configured to run on any compliant network with a five-second software update, or PC makers and carriers may agree deals to limit a PC to certain carrier networks in the more traditional way. To date, most embedded 3G has been carrier-specific and the result of one-off agreements between cellcos and vendors (such as Dell’s HSDPA offering with Vodafone in Europe).
At the recent Mobile World Congress in Barcelona, Qualcomm showed off a deal with notebook maker Hewlett Packard, which praised the way such products make it easier for PC vendors to put a 3G network card inside a laptop without going through multiple carrier certification programs. The Gobi-enabled HP computers could be available as early as June and the support of two big name PC makers is vital to indicate to the market that these companies, at least, see an incentive to devote precious space inside the laptop to 3G, even when they already offer Wi-Fi.
The PC makers’ willingness to give up real estate to 3G is not just a commercial decision, but also a performance one, since wireless is hard to embed in laptops because of the noise generated by the motherboard, and having Wi-Fi already in situ complicates the situation further.
Here, Qualcomm has managed to use its undoubted excellence in integration to provide a system that is pre-optimized to live in a notebook, and it will apply this expertise to other devices that could expand its reach beyond the handset. Even more than its integration skills, Qualcomm has offered the PC makers something that has often seen it through stormy waters ‘ its close relationship with carriers, which it has cultivated more assiduously than its rivals, partly because of the love-hate attitude of some of its vendor customers.
This has eased the process, according to Qualcomm’s Mike Concannon, VP of strategic products and, of getting Gobi pre-certified by all the major carriers, a feat that most chipmakers would be hard pressed to achieve in a short timescale. For users to reap the benefits, the main requirement is for the operators to be more creative. Qualcomm and HP, when showing off Gobi at MWC, called on cellcos to adopt a more flexible approach to data access with limited-time or volume connections ‘ once a carrier is selected at the splash page, Gobi can automatically load all that carrier’s settings from the laptop’s hard drive.
Qualcomm is certainly not the only company pushing the embedded 3G model. Ericsson is, predictably, a major supporter, and at MWC said it would partner with China’s Lenovo to embed HSPA in the PC maker’s ThinkPad notebooks, although while its Mobile Platforms unit stays ahead of the game in terms of putting the latest technologies into notebooks, it lags Qualcomm in the all-important commercial breakthrough of multicarrier, adaptive products. Notwithstanding, Ericsson predicts penetration of embedded 3G/3G+ in laptops will reach 50% by 2012.
Nokia has also been working on embedded modules, although its joint venture with Intel, focused on HSPA in the laptop, broke down last year on conflict of interest. This highlighted Intel’s dilemma sharply ‘ in theory, a combination of Nokia’s device dominance and Intel’s iron control of the PC value chain should have been unbeatable, as well as expanding Intel’s limited access to Nokia contracts, but it would have forced the chipmaker to accept that the technologies over which it has real influence, Wi-Fi and WiMAX, were just two options among many, and potentially under threat from 3G.
There are still criticisms of the embedded approach, and many operators currently remain more keen to offer USB dongles than get into complex agreements with PC makers or even have to subsidize laptops. Vodafone recently announced HSDPA wireless sticks, while 3 UK’s dongles come in various color choices.
Of course, PC makers will only pay decent money for an embedded module if they know it will boost the sales and/or premium price of their products, and that case has not been proven. But this has not been because of unwillingness to use laptop 3G, especially now, in many advanced markets, availability is greater than Wi-Fi’s and bandwidth just as acceptable for all but the most media intensive application. Rather, it has been because of the cost and complexity of roaming and charging.
This justifies the Gobi approach, which we are sure to see emulated by Qualcomm’s rivals ‘ with Nokia and Ericsson in the strongest position to negotiate with carriers and also source advanced chips, though Intel may still apply its own emerging adaptive radio technologies to the notebook where it rules supreme. The chipmakers will rise to the occasion ‘ now the onus is on the operators not to miss this opportunity to make mobile 3G data the default choice for their highest value subscribers, rather than an optional extra when there is no hotspot in reach.