Original URL: http://www.theregister.co.uk/2007/10/11/intel_nokia_mobile_internet/

Intel and Nokia make first step to forming mobile internet axis

Nokia to use Intel's chipset in WiMAX devices

By Wireless Watch

Posted in Networks, 11th October 2007 12:55 GMT

Comment The announcement that Nokia would use Intel chips in its forthcoming WiMAX devices was hardly unexpected, given that the two giants have been working together for more than two years in this area.

But it is still very valuable for Intel to have gained a firm contract rather than just a development partnership, since it has looked at risk of creating a market where, in the early stages at first, other chipmakers would reap the revenues.

More broadly, the Nokia win sets out the strategy that stands the best chance of reshaping the wireless industry in the internet image, with the chief representatives of the PC and the cellular heritage coming together to create a new breed of devices that will be optimised both for broadband web services and for mobility.

Nokia will use Intel's Baxter Peak chipset in its N Series Internet Tablets, the first WiMAX handheld Nokia plans to ship in 2008. Intel, Nokia and Nokia Siemens also announced an interoperability agreement that will ensure interoperability between Nokia Siemens' WiMAX and Nokia devices based on Intel chips.

The main WiMAX products from Intel are Rosedale 2, which is not fully wave 2 compliant but supports both fixed and mobile standards; Montevina, the next generation of Centrino Duo, which will offer integrated Wi-Fi and WiMAX for notebooks and UMPCs, and is supported initially by Lenovo, Acer, Asus, Panasonic and Toshiba; and Baxter Peak, a handset/PDA chipset.

Montevina incorporates the Echo Peak Wi-Fi/WiMAX chipset and also features integrated HD-DVD/Blu-ray support. As an interim solution, designed to seed the WiMAX market, Intel has also developed Dana Point, a single-mode WiMAX chipset that will be placed in PC cards.

Why Intel and Nokia should team up

It has always been logical that Nokia and Intel should cooperate to ensure that their vision of the mobile internet, which has many points in common, defines the next generation of communications. Their conflicts of interest are short term, while their mutual interests are long term and highly strategic.

Previous attempts to join forces, notably in HSDPA PC cards, stumbled on those conflicts – notably Nokia's continuing reliance on the cellco customers, whose business models are threatened by the flat rate, open access approach of the PC/IP world as represented by Intel; and the competition of the two companies to shape the mobile internet device itself, as enshrined in Nokia's Wi-Fi/Linux Internet Tablet and Intel's Ultramobile PC (UMPC) architecture.

On the first count, while the cellcos will remain Nokia's largest customer base for many years, it knows that in the open internet world, its power will be diminished. It will face competition from a wider range of operators with non-traditional models – potentially new clients for Nokia – and they will be forced to adapt to the open web anyway, as 3 and others are doing.

The faster these two developments take place, the better for Nokia, since in the open internet, the brand and functionality of the device hold sway, not those of the bit pipe carrier. Also, it will be able to accelerate its own moves to become a web services provider, with offerings like Music Store and Navteq that can bypass cellcos altogether.

All this puts Nokia in the Intel camp more firmly than ever before, especially in the US where it has failed to gain significant market share or carrier favours. And this could lead, in turn, to a rapprochement as the two join forces to define the mobile internet device architecture, rather than competing on this.

Intel will inevitably dominate the WiMAX laptop market, but in handsets and other devices its success is far less assured, and the support of a Nokia would be a turning point. Intel has failed in cellphones before – and last year sold its XScale operations to Marvell – and even in WiMAX, the specialist chip designers are steering clear of taking on the giant in notebooks, but are exploiting its inexperience in low power gadgets to come up with offerings that look, at this stage, more attractive than Intel's own.

Intel HAS to succeed in mobile devices - smartphones, media players, low cost widgets for low income communities and so on - or it will become stuck in laptop replacements, a growing market but not large enough to carry the future of such a giant.

Adopting WiMAX and Linux are good tactics, but remain heavily focused on the traditional PC customer base, some of which - Dell included - are struggling. A company of Intel's scale, to be credible in the new mobile generation, needs Nokia as a customer and an ally.

So Nokia's choice of the Intel chip for its first WiMAX tablets is a major credibility boost, however political it may have been, as well as a sign that the creation of an Intel-Nokia axis around WiMAX, which has looked logical ever since both stood together at the creation of the WiMAX Forum in 2003, is likely to come to pass.

This could hasten the development of a real volume market for mobile WiMAX handsets, but could cause problems for Texas Instruments and other traditional Nokia suppliers, and, as the Finnish company strengthens its web of alliances, for other contenders for the mobile internet throne such as Google, Microsoft, and the largest cellcos.

The challengers to Intel's WiMAX chips

The WiMAX CPE silicon specialists are trying to maximise their impact in the run-up to 802.16e wave 2, the key platform for mobility, and while they still have some respite from a full-on challenge from Intel, which may have got its story right, but has not yet executed the actual chips perfectly.

Wavesat, Beceem, Sequans, Runcom, GTC and others are all seeking differentiation in this market, and focusing on adding value to the base WiMAX specifications, in order to stay ahead of the sector, and of Intel. Also playing in this increasingly crowded space are Comsys, Altair, NextWave, RedDot Wireless, Redpine Signals, TeleCIS (in 802.16-2004) and Troicom.

Meanwhile, other majors are set to move in too. Motorola showed off its own 802.16e CPE chipset design at the recent WiMAX World event in Chicago and is also working with Texas Instruments on handset silicon.

Fujitsu has not had the impact that its early moves into WiMAX might have promised, but is ramping up its efforts and promising end-to-end products from CPE and base station chips to full systems, something also pledged by NextWave (which even throws spectrum licenses into the deal).

Samsung is likely to use a combination of its own and third party silicon for different ranges, and its closeness to Beceem, in which it is an investor, has made this start-up hotly favored to pick up some of this vital business.

Qualcomm, of course, remains officially hostile but owns various OFDMA assets, such as the former 802.l6e developments and team from TeleCIS, and could move quickly if it saw an unmissable market opportunity.

Wavesat and IBM

In their efforts to add value to the base specifications, the specialists are focusing heavily on critical issues that remain imperfectly solved in WiMAX devices, notably reducing power consumption, especially in multimode formats, and so integrating as much functionality as possible on to the chip.

Wavesat, whose main successes have been with Taiwanese ODMs, scored a notable coup by signing up IBM to manufacture and codevelop its Panther CPE silicon (official name UMobile 802.16e). The companies will use IBM's Power Architecture running at up to 400MHz in a design that can support functionality "beyond wave 2" in areas such as MIMO and on-chip memory, according to Wavesat president Vijay Dube. The deal may also signal IBM's intention to get serious about WiMAX as a market for Power.

In July, Singapore's Nex-G set out a roadmap including macro WiMAX and femtocell base stations and CPE, all using the IBM PowerPC chip, and said it would create a reference design based on the Power architecture. Nex-G is also a Wavesat partner in the fixed WiMAX arena and these various activities are likely to converge further.

The other Power major is Freescale, which has been less active than IBM in ploughing R&D funding into WiMAX partnerships, a factor that lost it an original deal with Nex-G.

For Wavesat, the key differentiator against Intel is putting 4Mb of memory on chip, doing away with the need for external memory except in the highest range devices. The first Panther-supported devices will be aimed at the Korean market and will be USB dongles and dual-mode WiMAX/CDMA PDAs and handsets, with SKT and consumer electronics companies being the primary targets (Wavesat already has a relationship with SKT).

The new chips will sample later this year, though this has been delayed from the initial deadline of September because, as Dube explained in an interview, customers are demanding more specific and non-standard functionality than he had originally anticipated.

He told WiMAX Vision: "For example, in Korea SK Telecom is requiring media independent handover (MIH) so end users can handover between the WiMAX network and the 3G network. Another example is customers' demands on the 624 QAM uplink modulation."

All this will make certification vital to ensure products remain standard, but will also require additional efforts at advanced interoperability testing, beyond the Forum profiles, among vendors and chip designers.

Other WiMAX CPE moves

Of course, the challenge of beating Intel will get tough once the giant starts to operate on all cylinders. The company said in its WiMAX World keynote that it would transition from "Intel Inside" to "Intel on the Internet", and that would involve making WiMAX chips "abysmally cheap, abysmally fast".

Copyright © 2007, Wireless Watch

Wireless Watch is published by Rethink Research, a London-based IT publishing and consulting firm. This weekly newsletter delivers in-depth analysis and market research of mobile and wireless for business. Subscription details are here.