Intel wields tiny chips to crack mobile devices

At least one half of the Wintel double act seems to getting some of its mojo back.

Intel, like Microsoft, made the mistake of looking at the world through PC-tinted spectacles for far too long. But the world’s biggest chipmaker by sales this week had a chance to show itself in the most favourable light, displaying its impressive ability to push computing to ever smaller and more efficient places.

Mastering a complex manufacturing process at massive scale has always been Intel’s core strength, and new chief executive Brian Krzanich seems very much at home. A typically pragmatic product of the chipmaker’s demanding manufacturing treadmill, Mr Krzanich – known universally as BK – looks comfortable at the helm of the silicon juggernaut. But he still has to solve the problem that bedevilled predecessor Paul Otellini, of how to use Intel’s unmatched assets to make it more than an also-ran in the booming mobile markets.

First, the good news. Intel is pressing relentlessly ahead in shrinking the elements in its chips to a scale where competitors will be stretched to match its levels of power efficiency and performance. That means not only the official launch this week of Quark mobile chips with elements measuring only 22 nanometres – or billionths of a metre – across, but a demonstration of 14 nanometre technology that will start to move into production next year.

Managing these cycles has been complicated by the need to protect a highly profitable PC quasi-monopoly while pushing into new areas where competition is fiercer and prices lower. Inevitably, sales of Intel’s low-priced, “good enough” Atom chips for the mobile markets threaten to eat into sales of more expensive Core line of products for PCs. The twin terrors of margin dilution and product cannibalisation start to rear their head.

The first of these concerns should be easier to lay to rest than the second. As the high volume producer, and a leader in packaging an ever wider array of functions on to its system-on-a-chip products, Intel should be capable of turning a decent margin at lower price points, provided it doesn’t slip up. This is clearly the bet the company is making as it lines up for what could turn into the next big opportunity for silicon: the plethora of smart devices that can be worn, embedded in everyday objects and even ingested, making up the twin markets of “wearables” and “the internet of things”.

The likelihood of cannibalisation is harder to handicap. All tech companies talk a good game when it comes to promising not to get trapped in a shrinking, high-price ghetto of their own making. Mr Krzanich has indicated he will not make the mistake of his predecessors in defending a shrinking PC turf to the exclusion of big new opportunities. We’ll see if his investors feel so sanguine if that means eating into PC profits.

Now the not so good news. Intel still has to prove that it really knows what its customers want – and that means not only the hardware makers who buy its chips, but the ultimate users who depend on them in their digital lives.

This is where its PC business model has come up short. Selling a single flavour of its main chips has starved hardware companies of the chance for differentiation. Offering more customisation has become a necessity.

Intel management claims to have got the message. However, diagnosing the problem on the one hand, and turning a manufacturing supertanker into a customer-led organisation on the other, are very different things. Mr Krzanich has his work cut out.

A second conundrum will be harder to solve. However impressive Intel’s chips, they are only as good as the devices in which they are shipped, or the digital services that they bring to life.

In this, it has faced the same problem as Microsoft, the other main platform technology company of the PC era. The software maker’s response has been to branch out into devices and services itself.

This is territory where Intel has never looked comfortable, as evidenced by its occasional ventures into consumer products in the past. But the coming of the internet of things may present a new opportunity to experiment. And, as its trials of a multi-channel TV service demonstrate, Intel has also been keen to master skills as varied as user interface design and digital service delivery.

Depending on how Mr Krzanich addresses issues like this, Intel on his watch could be facing an overhaul as big as the one that saw it move into microprocessors in the first place.

Richard Waters is the Financial Times’ West Coast managing editor

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't cut articles from and redistribute by email or post to the web.