With Intel considering a possible LTE/ WiMax tie-up, how will that influence the traditional dominant players of the cellphone market?
According to Richard Kramer at Arete, a research and analyst company tracking the wireless industry, only thre chipmakers shipped over 100 million units in the wireless industry last year. Those three were Nokia/ST, Qualcomm and Ericsson.
This burgeoning scale meant that others were rapidly losing out. Kramer asks, who is going to take LTE chipsets, and at which cost. He predicts that without support from a top tier OEM, such as Nokia gives ST, TI and Broadcom, and Ericsson gives TI, then other players are going to lose out.
Given the fact that LTE will be responsible for perhaps 6-85 of handset volume, and Ericsson and Nokia will want the lion's share of that between them, then can the other companies support the level of R&D investment required to keep up?
Kramer envisages a model where IPR costs passed down to operators have to lessen, as chipmakers follow a different model.
Certainly this is the route Ericsson and others took when, Alcatel-Lucent, Ericsson, NEC Corp, NextWave Wireless, Nokia, Nokia Siemens Networks and Sony Ericsson Mobile Communications all signed on to a press release stating a "mutual commitment to a framework for establishing predictable and more transparent maximum aggregate costs for licensing intellectual property rights that relate to 3GPP Long Term Evolution and Service Architecture evolution standards."
The parties said the framework is based on fair, reasonable and non-discriminatory (FRAND) licensing terms. The companies noted they support a maximum aggregate royalty level for essential IPR in handsets as a single-digit percentage of the sales price. For notebooks that might integrate LTE modems, the companies agreed to a single-digit dollar amount as the maximum royalty level.
That level of IPR take would probably mean that the vendors would seek profits instead through a channel pricing model, rather than on steep IPRs. This could and should make the cost of introducing a new technology more acceptable to operators who faced delays and cost with the development of 3G, where handset development was held back by disputes over IPRs and the costs of IPR.
Yet as well as Kramer's predictions, there is another area of investment plunging into OFDM chipsets that also threatens the business case of many players in the semi market -that of the WiMax chip vendors, and specificall Intel.
Intel sees itself as a company that can create markets b the force of its own investment – witness WiFi and now WiMax. It also sets trends, in order to seed its own future revenue streams. The problem for Intel as devices go multi-mode, and service offerings converge, is that it is nowhere in cellular chipset market. Of course, there is also an opportunity, as it could lead convergence from the PC/ WiFi side of the fence.
There was a notable softening of Intel's "WiMax is best" stance (to greatly simplify their arguments about time to market, cost and technical superiority)recently when Sean Maloney joined the small group of industry heavyweights calling for the wireless industry to unify its OFDM paths, which currently lie separate as LTE and WiMax.
Vodafone's soon-to-be-ex CEO Arun Sarin famously mused that perhaps WiMax could form the TDD part of LTE, but many thought Intell, which stands to gain the most from a vibrant standalone WiMax market, would resist.
Yet other chip makers in the WiMax market are quite happy to countenance the idea not just of making chips for both LTE and WiMax developers, but of developing multi-mode chips that would support devices to operate either on both standards, or be "switched on" to one or other standard as the OEM required. If you add to this the difficulties with the patent pool within WiMax, and the slow perdio of standardization and market development, a warmer Intel attitude toards LTE makes more sense.
Speaking to the BBC, Maloney said, "In our view they ought to be harmonized," He agreed with the industry consensus that when it comes to the iair interface, the two camps are about 80% similar.
"The main difference is that WiMAX is a couple of years ahead," he added, interestingly.
Apart from this being the standard WiMax camp line, this is important because wit so much development and R&D dollars sunk into WiMax already, there might be a case to be made that the cheapest way to get to LTE chipsets would be to leverage much of that investment (and associated IPR, obviously).
Maloney pointed out that is was possible even that the Wi-Fi/WiMAX Centrino 2 could be adapted for LTE, though he highlighted this"would certainly be a nice long term goal".
Just as interestingly, given our boxout on the MID market, Intel hopes that its Atom mobile internet device chip will give it with a re-entry into the mainstream mobile market after the sale of its XScale mobile processor division to Marvell. Atom's mobile form, known under the code Pineview is said to be ready in 2009 – a year before the first LTE networks are expected to be commercially live.
So there are signs that the traditional cellular chip vendors and developers may not have everything their own way – there are their own internal challenges to be addressed – and those of their ultimate end customers, the operators. Then there is the challenge from the WiMax developers, and there are players here such as Sequans who have valuable businesses – it's not all about Intel.
But crucially, the cellular players will view this as the market coming to them – not them having to work out the threat of Intel.
"There's a lot more to this business than making chipsets to run in PCs and notebooks that are only ever really used in a static environment. We understand the demands of mobility, of power management, of processor power, that mobility paces on device," one developer told us. "Intel has made a great business with WiFi, and is now having less success with WiMax. It realizes the future is converged, mobile devices, and it is sitting on the wrong side of that convergence divide."
However the market breaks down over LTE chipsets, there is perhaps a bigger question to ask at the moment. As operators sweat 3G and HSPA assets or some time to come, will there be enough end investment in LTE to justify any of these business models. Now that really is a bearish question.
Looking Forward to the MID range
Forward Concepts has announced the publication of a new study that covers the market prospects of the emerging Mobile Internet Device (MID) that serves a gap between high-end smartphones and ultra-portable PCs (UMPCs).
Mobile Internet Devices (MIDs) represent a new class of mobile communications and lifestyle devices. Its hardware, software and form factor will require design from the ground up in order to meet market requirements for features, price, performance, and power requirements.
– The user interface will be key to success and will likely need to be capable of responding not only to touch-based inputs but also keep pace with other evolving input methods such as ones based on motion, gesture, placement, etc.
– Although Apple's 3G iPhone plows new ground in Internet access, user interaction and utility, we don't consider it to be a MID, since we believe a true MID also requires a larger (4- to 6-inch) screen with higher resolution (VGA), TV out and optional Mobile TV capabilities.
– Global MID shipments will grow from 305,000 units shipped in 2008 to almost 40 million in 2012, reaching $12 billion in revenue.
– Integrated circuits for MIDs are forecast to grow from $29 million in 2008 to $2.6 billion in 2012.
– Intel has a much better shot at UMPCs, being predominantly an enterprise play, where x86 compatibility is important, and with battery life expectations in line with notebooks.
– Texas Instruments, with its mature and proven OMAP application processor family and the largest market share of the stand-alone Smartphone applications processor market, is one of the two best-positioned non-X86 semiconductor vendors for supplying stand-alone applications processors for all classes of MIDs.
– Qualcomm is the other best-positioned non-X86 semiconductor vendor, with its powerful new SnapDragon application processor and the company's market-leading 3G wireless solutions required for the MID market.
– Other chip suppliers will have plays in the MID market, including Nvidia, with its strong graphics capability which will play well for gaming applications and Samsung, with its applications processor experience and stacked memory capability.