As buses drew up outside the Consumer Electronics Show, and 140,000 potential customers disembarked, Intel began working the crowd. What were they pushing? Television.
For the world's largest chipmaker, the PC has become something of a straitjacket. Notwithstanding Intel's success in building and maintaining the market, analysts keep asking, 'Yes, but have you got anything else?' The company has sought to oblige such requests with forays beyond its principal microprocessor business - but it has met with little success.
Founder Gordon Moore still jokes today about his $15m digital watch - one of Intel's early consumer electronics flops. It has amassed further failures, mediocre or stalled ventures in sectors like mobile communications. The corporate irritation has been easy to see, even as the MPU juggernaut has advanced.
But are things about to change? An evolution could be at hand that may allow the company to position itself as the start-to-finish technology provider for the next wave of social digitisation. It all revolves around the 'humble' television.
Intel staff were working the crowd outside last month's Consumer Electronics Show (CES) in Las Vegas before anyone even got close to the company's stand, handing out copies of 'Screen Future' by company futurist Brian David Johnson.
Johnson's thesis is that the TV is about to once more become the technological centre of the home. It will, however, be much changed, as it incorporates more processing power and more applications and functionality. Consider this four-step progression: Informative TV to Ubiquitous TV to Personalised TV to Social TV.
Intel's stand still resides directly opposite Microsoft's, a historical reflection of the 'Wintel' alliance. But where the displays were once tightly coordinated, they now offer divergence. A CES newbie could have traversed both and come away with little idea of the position either company can claim in traditional computing.
This year, Intel's stand focused on 'a new vision for television'. This may sound like a delicate distinction, yet it is also an important one: there were plenty of laptops connected to 50in displays, but not so much to allow the permanent throng to see what was on the smaller screen and to prove what those laptops could deliver to a living-room.
Meanwhile, 'Intel-inside' kit in pride of place was more likely to be the Freedom Revolution, an IPTV set-top box/Blu-ray disc player from French operator Iliad, than another iPad wannabe tablet. The Revolution, with a sleek look from iconic designer Philippe Starck, is not just a cool design win for Intel. Iliad's Free subsidiary got commitments of two million boxes within hours of taking orders last December, making it a big, commercially significant win in the TV space. And there's more to this than a bid to simply recreate AppleTV.
Apple has set the agenda when it comes to marrying computational power to content consumption. The iPod, iPhone, iPad and AppleTV are all about delivering media in the appropriate location over the appropriate network in the appropriate format via the appropriate device. In each case, traditional PC 'productivity' is downplayed, leaving the market open to devices like those from Research In Motion's BlackBerry family. Apple has mum, dad and teenage offspring devices.
Intel carries a bit of a grudge here. It provides MPUs for Apple's MacBooks and iMacs. But it was evicted from the latest version of the AppleTV (see Teardown) in favour of an ARM-based processor developed around Samsung's Hummingbird platform.
Replacing the PC
Intel is aggressively pursuing slots in rivals to Apple's range, but the vision goes beyond that. First, there is the desktop PC market. Research consistently shows home use transferring to portable devices. Desktops have their place in businesses, where the refresh cycle for hardware is consistent and mature. But, in the home, they are being replaced.
In 'Screen Future', Johnson quotes Jeffrey Cole, director of the Centre for the Digital Future at the University of Southern California's Annenberg School for Communication, on how 'competition' between PCs and TVs has changed from when the two competed for your attention.
'Generally your computer was located in a room that didn't have a TV,' Cole says. 'It was [in] the den or the kitchen. There are, of course, a number of technical and infrastructure reasons for this. It was where your phone was located and many people didn't have phones in the living room. So people left the TV and went online for 30 minutes, got their things done, and came back to the TV.'
These were the days of dial-up. Add broadband and wireless, and people can dip in and out of the Web from the sofa while the TV is on in the background. 'The Internet was an enemy of television in the dial-up era, and it is now becoming the best friend television has ever had,' Cole says.
Today's TVs incorporate a wide range of applications. They are connected to the Internet via home networks and provide access to YouTube, Facebook and Twitter as well as video-on-demand. Cisco Systems used CES to launch a consumer version of its commercial Telepresence videoconferencing hardware, so you can chat with friends while watching 'Doctor Who', follow tweets and get alerts if a new email arrives, all on the same screen.
Johnson's argument is that, while this is not 'traditional' TV, it is basically TV in a multi-faceted form, one we must start getting used to. 'TV or the experience of watching TV is moving off the wall, moving from that flat screen or the traditional TV in your media cabinet and spreading across a wide variety and number of consumer electronics devices,' he writes. 'So now when we look at the good old-fashioned TV, we're really looking at a dynamic entertainment experience distributed across multiple devices, locations and consumers.'
By 2015, there will be about 500 billion hours of content available online and some 15 billion devices capable of connecting to the Internet. That is more content than anyone can realistically manage and it will be accessible by more devices than twice the number of people on the planet.
As Australia's Minister for Broadband, Communications and the Digital Economy, Stephen Conroy, is quoted as saying in 'Screen Future', here is the critical consequence of that: 'When you've got a million things you can choose from, you need someone to aggregate for you.'
That is the role that Google has sought, using algorithms and massive amounts of storage to crawl the Web, make some sense of it and deliver search results to the user in less than a second. But the task of turning TV images into not just a digital signal but a similarly analysable datastream is set to take a whole lot more processing muscle. That is exactly what Intel specialises in.
The second-generation iCore chips that Intel launched at CES offered a clue as to where all this is leading. They incorporate graphics as well as features such as Insider, Quick Sync Video and Wireless Display to grab and distribute content up to 1080p.
A four-minute HD video file can be converted for use on an iPod in 16 seconds, implying six minutes for a 90-minute movie. With up to six cores, the speed on the processors can be, the company claims, taken up to 3.7GHz using a 'turbo' overclocking technology.
That places a massive chunk of processing power locally for consumers to shoot and enjoy content around the home. Beyond that, Johnson describes a hybrid network architecture that does a lot of the crunching elsewhere on powerful servers, another key Intel market.
What if a system could recognise that a clip features a certain type of content and then build applications and services around that? The company's Beijing lab has developed a system to analyse a football game, recognise the players, what they do in the match, what goals are scored and so on. Imagine that capability set to recognising actors in a film, politicians or celebrities in a news clip or even which members of your family appear in different photos.
In commercial terms, what if you can use that analysis to propose and aggregate content according to how you profile the user - perhaps even look from the screen to see who specifically in the family is in front of the TV and then refine that process still further? For a start, you can begin to make sense of those billions of hours of content for the consumer, but you can then also start to target other information: apps that run in parallel and, yes, advertising.
Intel has that technology in its labs right now and, perhaps more important for a company known for keeping its secrets, is willing to let Johnson write about it and discuss in more depth.
The 'TV everywhere' model that Intel is adopting does, therefore, go beyond what Apple has built and brilliantly exploited. For a processor company, the approach clearly reflects the evolution of consumer end-usage, as it shifts from a general-purpose PC to a plethora of devices but with the largest home screen as the hub, a local head-end.
Going further, there is a clear bid here to use the company's existing strength in industrial-level computing to dominate the backbone - and then wrap everything together within one consistent infrastructure. It may seem a natural step, but it is an almighty bet.
That said, Intel's position in traditional microprocessors and servers gives it the foundations from which it is probably the only company that could gamble so aggressively for this market.
This also points to another significant difference between Intel's pitch for tomorrow's TV market and its other forays into businesses beyond the PC. This one is integrated with - as opposed to allied and parallel to - its existing strengths. It also looks like an end-run around the explicitly consumer-facing focus of most other content consumption and distribution strategies, although it does include it.
By attacking the TV ecosystem holistically, Intel now has something it can sell to consumers and also content providers and advertisers, panicked as they are by what this new type of television is doing to their revenues. The need for companies to 'monetise content' is well recognised, and Intel has one of the more plausible technological strategies out there.
Although there is another side to that execution. Speaking at 'Storage Visions' in Las Vegas, Johnson noted that if you see a PC stop working in a movie, a bug is blamed and the user simply reboots. But if you see a TV with no picture, it means some alien bugs have landed and they aren't friendly. Unfortunately for him, those comments came before Intel had to admit to a flaw in the chipset surrounding those new muscular iCore processors.
Nevertheless, Intel is fixing the problem. More pertinent to its longer-term goals, however, is what could emerge at this month's International Solid State Circuits Conference, where the company is scheduled to present its latest server thinking in a paper on 'A 32nm 3.1bn-transistor 12-wide-issue Itanium processor for mission-critical servers'. After the set-top box, the infrastructure.