<font color=Orange>Rethink on Convergence
A MUST READ!</font>
<font color=DarkSlateGray>
[The other giant involved in WiMAX is Nokia, pointing to an increasingly strong partnership between Intel and the Finnish handset giant, which could rival Wintel in its impact on the technology market.]</font>
<font color=RoyalBlue>Intel, a company that wants it all
</font>
Intel may not be the world’s largest chipmaker, but it is the most influential, and seeks a role in defining the IT industry far beyond the usual expectations of the silicon brigade. With moves into wireless it is looking to shape the next generation of computing.
Published: Sunday 1 June, 2003
Intel is a schizophrenic company. It has a core business in the guts of the computer, where vendors normally remain anonymous to the non-specialist world, yet its dearest goal is to have one of the world’s most recognized brands. It has succeeded in creating the only chips that are household names in their own right with Pentium, with the end in sight not just corporate vanity and shareholder value, but utter domination of the PC market.
Intel wants it all. It wants a Hoover-class brand, where its product name is synonymous with the generic product. But Hoover actually sells only a small proportion of the world’s vacuum cleaners. Intel also wants to be SYG, the company that makes dust filters for almost 100% of the world’s vacuum cleaners. And if it can achieve both sides of this coin, it wants to ensure it uses its power to drive the development of computing in directions that play to its strengths, that will require ever more of its chips and where it will be hard for competitors to challenge.
This is a hugely ambitious strategy, especially in this critical year for Intel. Like its witch’s familiar Microsoft, the chip giant faces not just depressed IT markets, but the decline in ubiquity of the PC as a computing platform, and a dangerous crossroads in its fortunes. Both companies need now to play on platforms where their domination is not guaranteed, going upmarket into the server back end and outwards to the mobile client. Intel will throw its usual powerful weapons at these markets – its financial might, which funds its innovativeness; its incomparable influence over PC manufacturers; its incongruous expertise in branding. It recognizes, as no other component maker does, how to work alliances and established sources of strength in order not just to achieve a strong position, but to help set the goalposts itself.
This is why Intel is of relevance to a CIO publication, rather than being confined like most of its rivals to the pages of the electronics press. Why should IT decision makers really care? In general, they don’t. The Intel branding campaigns are aimed primarily at consumers and personal business users and the CIO is concerned about what the computer does, not who powers it. But Intel’s position as one of a handful of companies that actively influences the direction and pace of development in enterprise technology makes it a company of critical interest to the IT strategist. This is a company that wants to mould the business (and consumer) IT market, not to be just the engine maker, and will influence the technology choices available to the CIO over the coming years. If it succeeds in its goals, we will have one picture of computing; if it fails, we will have another.
Why does Intel hold such an influential position?
It is not the largest chipmaker in the world.
It doesn’t have the strongest brand in the world. Although the ‘Intel Inside’ initiative has gone down in business school history as one of the most successful rebranding exercises of all time, Intel is not quite up there with Hoover, Biro and other names that are synonymous with their product type (This, incidentally, is a failure it hopes to put right with its Centrino wireless chipset.)
It is certainly not a consistent success story. It has had its share of financial tribulations (see finances box) and its fight to establish Pentium as a respected server processor to play alongside the big Risc boys such as Sparc and PowerPC has been long and bloody. In the enterprise it has been hampered by its inextricable association with desktop PCs, which make it appear a lightweight choice for the large server, and by its marriage to Microsoft, whose operating system is also struggling for enterprise respectability on a broad scale. And periodically it faces challenges from its own clonemakers, most recently AMD with its Opteron server processor.
But for all these weaknesses, Intel is in a position, if it makes good decisions in the next few years, to help shape the technology future and even to increase its influence over the PC manufacturers, extending this to the new hardware powerhouses, the smartphone makers. Its overall aims are not hard to understand – to support and drive any technology change that will increase the requirement for its processors and other chips; to develop products and relationships that can take a number one position in as many sectors as possible including enterprise servers and mobile clients; and to use its market weight and its ability to innovate to become entirely synonymous with personal computing of all natures. Intel aims to be the only sane choice for the manufacturers of PCs, PDAs, cellphones and future devices.
That means being unfaithful to Microsoft, unless the software giant can turn around its mobile fortunes and make Windows dominant on the new platform. Intel is putting itself in a position not to care about the outcome of that particular battle. If Wintel fails against Symbian in the most rapidly growing market in enterprise IT, Intel not only has a credible survival strategy, but will find all kinds of benefits to being less closely bound to Microsoft.
Intel’s public image has transformed almost overnight from that of a chipmaker with wireless interests to a wireless operation – notwithstanding the fact that its first quarter results were dragged down by its wireless business, where revenues dropped 29%. CEO Craig Barratt shrugged this off, saying a seasonal drop in wireless sales was expected. Intel is not going to weep about its $473m wireless sales last quarter, when it is focused on far larger prizes – the market share it believes it can achieve for Centrino and the Manitoba/XScale cellphone chipset, and the opportunity to set the goalposts in the most important emerging technology sector. Wireless is currently a very small percentage of the giant’s revenues but accounts for about one-third of its R&D.
Nowhere is this more true than in the mobile and wireless sector. If you just read the headlines and the famously expensive advertising campaign, you would think Intel’s mobile strategy rests on Centrino. But Centrino is just the tip of the iceberg. It is important because it illustrates Intel’s approach to brand and the attempt to create a brand that will drive customers to demand certain innards to their computers. Intel Inside was one stage but Centrino is even more aggressive. Intel knows that the provider of the hidden insides of a product is, to some extent, disposable until the customer actually believes that they will get a better product by demanding certain components. It’s a rare feat to pull off in any industry, but one that gives enormous power over the manufacturers.
It also carries risks, notably that of alienating the chipset makers. These have traditionally worked reasonably well alongside Intel, but by stating that PC makers can use the Centrino brand only if they have the entire package of the Pentium M processor plus the network and wireless components, Intel is taking a new confrontational stand against the makers of the lower cost third party chipsets that surround Pentium, in the name of extending its brand strategy beyond Pentium itself.
But technologicially Centrino is offering nothing very exciting. Intel is making Centrino its second most important business focus after Pentium because it is gambling that demand for wireless and mobile functionality will drive flagging demand for new PCs. But it is not creating the wireless market trend, and though it will certainly accelerate it, it has limited influence over its speed and development.
Even CEO Barrett, launching the chip, admitted Intel was not doing anything revolutionary, but was offering technology that would make wireless more usable. That is certainly true. Centrino will enable PCs that can run wirelessly and without a power cord for up to six hours (in the case of the upcoming new Sony Vaio models), though more commonly five or less. That makes Wi-Fi a more practical proposition for serious business usage on the move.
Nobody questions Intel’s ability to accelerate the uptake of Wi-Fi-enabled PCs. Already 90% of the world’s PC makers are committed to adopt Centrino and the marketing campaign is sufficiently aggressive to expand the edges of the consumer wireless market significantly. But for now, most people are still looking for a network, and without one, buying a Centrino PC is like having a DVD disk but no player. Intel cannot hope to speed up the hesitant corporate adoption of WLans significantly – full standardization of the faster and more secure 802.11a and 802.11g technologies will do that, as well as falling costs.
Its real hopes are invested in public access WLans. It believes it can influence the spread of these facilities, not only by investing in hotspot operators, but also by creating, with Centrino, such a climate of user demand that roll-out of hotspots will be accelerated. Whether or not this succeeds, Centrino is accelerating the market, not adding anything new, especially in the corporate space.
That isn’t true of WiMAX and the personal data server, however. With these innovations Intel is seeking to shape the market in its own image and to take a dominant position in any approach to mobile computing that gains favor with the enterprise community, rather than putting too many eggs into the Wi-Fi basket.
The personal data server could, if successful, kill the PDA and change the nature of wireless computing. Roy Want, the principal engineer at Intel Research, is developing a device half the size of a PDA that holds the owner’s data but relies for its display and keypad on computers or smartphones that happen to be within range, linking to them via Wi-Fi or, more probably, a short wave technology such as Bluetooth. Eventually, Intel envisages it being integrated into the smartphone, something that could make its Xscale processors, on which the prototype server runs, as dominant in cellphones as Pentium is in PCs.
The development is part of Intel’s year-old Ubiquity Project. The company has demonstrated prototypes recently and Want said that products could be available within three years, though that timescale was dependent on leaping some technical hurdles such as development of sufficiently high density storage – something Intel itself is working on - the establishment of standards in wireless short range communications, and improvements in power technology to allow the device to be always-on but to need recharging only once a week or so. Wireless support in PCs also needs to be ubiquitous. All these conditions for success benefit Intel, driving adoption of wireless technologies such as Centrino, its own current and future storage and smartphone processor designs, and protocols such as Bluetooth that is is helping to shape and support.
The project is the most significant of several vendor initiatives to explore a new pattern of computing in which data and applications are locally held and fully mobile, downloading wirelessly from central servers. But it goes a step further by preserving the involvement of the conventional PC, albeit with a downgraded role, and with wireless support a must. Such a picture – the wireless PC preserved, and the cellphone incorporating the Xscale-based personal server – is a dream for Intel. As Barrett puts it: “When we talked about the extended PC two years ago, we were just starting a big change in the industry. Now as we start to talk about broadband and wireless capability, I think we're even increasing the importance of the PC in the equation.”
WiMAX has a similar potential to drive demand for more powerful, functional PCs, laptops and cellphones, offering the prospect of Wi-Fi style low cost cards delivering long distance broadband wireless communications. The WiMAX initiative is a group of nine vendors promoting and developing the hitherto neglected IEEE 802.16 standard for wireless metropolitan area networks. The initial version operates in the 10-66GHz frequency band and requires line of sight towers, but the 802.16a extension, ratified in January, uses the lower frequency of 2-11GHz, easing regulatory issues, and does not require line of sight. The upcoming ‘b’ extension will add Quality of Service (QoS) standards to guarantee call quality, and will increase the spectrum to 5-6GHz. Most significantly, the mobile version, ‘e’, is likely to be ratified by the end of the year and could eventually sideline Wi-Fi, once frequency licensing and battery issues are sorted out.
The extended 802.16 standard is chiefly important – at least before the mobile version comes along – to connect Wi-Fi hotspots to the internet and backbones, and to provide a low cost wireless extension to cable and DSL broadband services. It boasts a 31 mile range compared to Wi-Fi’s 200-300 yards, and 70Mbps data transfer rates. WiMAX claims base stations will cost under $20,000 and support 60 enterprise customers with T1-class connections. It is likely that Intel will include WMan chips alongside, or as part of, Centrino within a year although it has announced no firm plans and claims its interest at present is primarily in driving broadband uptake in order to encourage PC upgrades.
The other giant involved in WiMAX is Nokia, pointing to an increasingly strong partnership between Intel and the Finnish handset giant, which could rival Wintel in its impact on the technology market. Nokia has many choices of its own to make – which long distance technologies to support, given the problems besetting 3G, but the hostility of its carrier partners to backing WiMAX as an alternative; how best to carve out a position in the enterprise market to ensure that its Symbian-based smartphones become the corporate mobile platform of choice; among others.
Intel aims to be neutral in such decisions, providing core chips for any approach to mobile computing and allying with the strongest partners in each sector. It will have Centrino and a future WiMAX chipset for its existing PC partners; and the XScale architecture for its new cellphone friends, a product that it hopes to push into the front ranks by offering the handset makers the carrots of the personal data server and the mobile WiMAX device. This is a roadmap currently unmatched by the established cellphone processor giants, notably Motorola, whose own personal data server concept is less ambitious and which has so far been uninvolved in WiMAX (as has Microsoft).
Intel is working hard to get an influential position in the smartphone market, which has a natural wariness of Microsoft and an affinity with Java and Symbian. The recent news that Intel is working with operator MMO2 to integrate their smartphone development communities – Intel’s PCA Developer Network and MMO2’s SourceO2 program – shows the chipmaker changing the rules to achieve new prominence for XScale. Like its main rivals in this space, TI and Motorola, it has traditionally delivered silicon to the handset makers and then had not further input to the process. But with MMO2, Intel will collaborate to fine tune XScale-based handsets such as the XDA, adding new layers of software to ensure optimal performance of key applications and services on the device. The partners will identify applications that have strong revenue potential, optimize them and market them through O2’s Revolution program.
At this crossroads in enterprise computing, when the world is poised to go mobile and the PC’s role is set to change, Intel is so far reacting like an old tiger that has glimpsed new territories and is in eager pursuit of conquests. By contrast, Microsoft is behaving like a defensive old lion whose core territory is being threatened. But all that could change. It is far too early to write off Microsoft’s so-far unimpressive attempts to formulate a mobile Windows strategy with CE and Windows Powered Smartphone. Just as it is far too early to say that Intel’s first impressive moves towards the new technology picture will assure it success. There are many other powerful forces seeking to achieve dominance of the silicon underlying mobile, some of them either with their own handset operations – Motorola for instance – or with stronger links than Intel with the cellphone heavyweights – Texas Instruments springs to mind.
While Intel gambles on the enterprise and the consumer going mobile, and on its ability to accelerate and dominate that process at the chip level, it must protect its own turf. The PC is not going away in the near future, and the enterprise server will only have its role enhanced by a move towards mobile clients and remote access, but Intel is fighting the global downturn in computer sales as well as various traditional rivals.
In servers, Intel has to fight against all its established Risc competitors, and now faces its first serious challenge in years from the company that made its name from cloning Intel processors, AMD. AMD has come out with a hybrid 32-bit/64-bit processor called Opteron, which sits halfway between Intel’s 32-bit Xeon and high end Itanium, but preserving software investment in 32-bit architectures. Intel has taken the stance that the world is ready for 64-bit computing and it wants to push every server supplier to make that same leap of faith and upgrade to Itanium. AMD is offering them a more cautious two-step migration, and its hybrid architecture is now supported by both Sun and IBM, runs Linux up to eight processors and will run Windows Server 2003 by the start of next year, with the possibility of 32-processor configurations – and all this at Xeon pricing levels. Intel faces a real danger of conceding the volume server processor market, currently dominated by Xeon, to AMD and having to pick its main battlefield at the higher Itanium end where alternative architectures remain strong.
This is just one example of Intel’s vulnerabilities as it seeks to entirely refocus its business on what it perceives to be the main growth areas for silicon. Its strategy is convincing, but what of its timing? Intel seems to have put its finest minds into extending its kingdom out into the mobile space, creating an even larger playground than it enjoys with the Wintel architecture. In doing so, it must be wary of leaping too soon into the mobile world, and endangering its strengths in traditional markets.
rethinkresearch.biz |
