Business Management

Intel shrinks jobs but does it regret its colocation U-turn?

Intel’s plan to cut 11 per cent of jobs, laying off about 12,000 people, underlines the fact that this is a company crunching through the gears of change. Reading the press release announcing the bad tidings, I was reminded again of a path that Intel took at the turn of the century before quitting it, and I wondered what would have happened if it had stayed on that track.

Some context: Intel is the world’s largest maker of microprocessors and made much of its fortune by providing the central processing units that act as the brains of personal computers. Every time we bought a Dell, HP, Compaq, IBM or other PC, usually running Windows, Intel was usually the source of the single most important component. For decades now it has been the dominant majority provider of these chips and in 2005 was even able to win over the one big stand-out in desktop and mobile personal computers, Apple.

As if that weren’t enough, just like Microsoft, Intel was able to segue into the market for volume servers, eventually also leading that segment too, having usurped companies like IBM, Sun, IBM, HP and MIPS. That powerful dual position led to vast riches and Intel remains a huge company with a market cap of about $150bn.

But Intel is facing up to some profound challenges. One of them is that many of us today use tablets, smartphones or other devices and those devices need tiny, low-power, low-cost chips. So far, despite several attempts, Intel has not been hugely successful in those markets, unlike ARM, the UK company that supplies cores to chip-making partners. Meanwhile, the PC sector is suffering a precipitous declines.

Worse for Intel, ARM is collecting more and more partners in volume servers too as buyers seek power-efficient boxes for their datacentres. ARM hopes to have a quarter of server shipments by 2020.

Another, related battle for Intel is in the general shift in the industry towards cloud computing. In this relatively new world many of the big providers look for the lowest-cost providers and even have enough sway to order customised parts. That is good for volume but potentially ruinous for margins.

Intel’s strategy for addressing these big changes has been both to compete head-on and to diversify, for example by buying McAfee in security software, investing heavily in the Big Data software leader, Cloudera, and betting on growth markets like wearables, the Internet of Things, programmable chips and, irony of ironies, the chips that will power cloud datacentres. But I wonder if Intel executives ever think about a strategy from way back circa 1999 when the company wanted to make datacentre colocation a big part of its business.

Back then, cloud computing was not a popular term but people talked a lot about application service providers (companies like Salesforce.com and NetSuite were just emerging at the time) and there was an increasing interest in datacentre hosting. Intel made a big splash in the market with a unit called Intel Online Services, spending a reported $1bn on the exercise with the hope that providing third-party data services would become a major plank in its overall business. The unit was led for a time by Renee James, an executive later to be linked with the CEO role at Intel.

The project began just as dotcom mania was reaching its zenith but when that market crashed in 2001 Intel lost faith and it eventually pulled the plug on the project in 2002.

Dalibor Vrsalovic, president of Intel Online Services, said this:

“While IOS has been successful in attracting new customers, market trends and financial projections for the hosting services industry lead us to today's decision."

If only Intel had known what would follow, with businesses moving to the cloud and a new crop of internet startups leading a business boom. If Intel had held its nerve it could have netted quite a catch. Today, the leader in colocation, Equinix, has a market cap of over $22bn. That could, and perhaps should - given the company’s influence on the market, brand and wealth - have been Intel’s money. Intel might even have been far-sighted enough to build platform software and challenged Amazon Web Services, a company that some pundits believe might already be worth more than Intel.

In a press Q&A I once asked the former Intel executive Pat Gelsinger if he regretted Intel’s exiting of the colocation market. He swatted away that question saying abruptly that it would not have been a good idea to persist with, but I know other executives of the company that privately disagree.

Intel remains a vibrant company but it must, in reflective moments perhaps, regret the road not taken or rather the U-turn that could might have led to a very different company.


« Why Africa needs a space program


OpenText picks over once-feted Autonomy software »
Martin Veitch

Martin Veitch is Contributing Editor for IDG Connect

  • twt
  • twt
  • Mail


Do you think your smartphone is making you a workaholic?