Column Intel has been many things. It has been a struggling startup, surfing the sudden successes and near-death experiences of 1970s Silicon Valley. It's been a relentless powerhouse, ruthlessly exploiting its dominance as personal computing rode Moore's Law to commercial and social revolution. Now it's in danger of seeing its supremacy crumble as rivals grab markets and mindshare with better manufacturing and marketing as Moore's Law itself turns to sand.
Chipzilla has also been gloriously unsuccessful. Mobile, graphics, wireless, display tech, optical, VLIW, RISC, a slew of memory and storage technologies – all have come and gone in the company's search to find a new world beyond full-fat x86.
Many were research projects that Intel hoped but did not particularly expect to catch fire, and there's no dishonour there. Some, like Itanium and Atom, were sent out fully armoured to do battle but just didn't find the right lands to conquer.
Some, like Intel's permanent inability to build a GPU that matters, remain mysterious. As long as the company could carry on leading with ever-smaller architectures and ever-yieldy fabs, it had the luxury of failure. Now it does not.
But Intel has had one attribute outside its control: luck. It was lucky when Busicom approved the idea of a CPU to drive its calculator design, and the company decided – reluctantly – to properly productise the resultant 4004, despite being a memory outfit first and foremost.
Pat Gelsinger first entered my life in 1988, although neither of us realised it. That 1980 decision by IBM had by then led – via the 80286, another Intel what-were-they-thinking misstep – to the 80386 chip, the first Intel CPU with an architecture that didn't ruin your appetite.
Gelsinger couldn't be more Intel if he had segment registers. (He may have segment registers. I'm not his doctor.)
As a young programmer at a startup, I was given the job of writing an 80386-based networking server kernel, for which I was grossly unqualified. My salvation came in the form of Programming the 80386, a book co-authored by one Patrick P Gelsinger. Although clearly talented as a technical author, he abandoned that path to lead the design team for the 80486, which was a modest success in the 1990s. Modest like Nirvana.
Around 10 years later, I first met Gelsinger at an Intel Developer Forum, and renewed the acquaintance once or twice a year until he shockingly left for EMC following what may be most kindly described as short-sighted decisions by the board.
At those many IDFs, I found Intel's long tradition of respecting and promoting engineers had led to a unique mix of personalities at the top, one where genuine enthusiasm for technology could still flower amid the corporate Dungeons and Dragons game. As Intel CTO, Gelsinger's engineering nous had to be and was sans pareil, of the sort fired by genuine engagement with possibilities. That he was one of the most transparently decent humans in tech did not, remarkably, hold him back.
And now Intel is astonishingly lucky in getting him again, burnished with the sort of tech leadership record at EMC and VMWare to which "right man, right time" barely does justice. The cloud is shot through with VMWare's stack, which speaks of a sustained, sure-footed strategy that engages with possibilities matched to needs.
But that decade of success at the top is built on 30 years of living and breathing Intel culture: Gelsinger couldn't be more Intel if he had segment registers. (He may have segment registers. I'm not his doctor.)
Which means now we're the lucky ones. We're about to get a ringside seat as Intel faces the existential question that all huge, established companies in a rapidly evolving environment must face – how much to change, how much to stay the same?
VMware has the strategy and culture to thrive after CEO Pat Gelsinger's exit to IntelREAD MORE
On the face of it, by anointing the quintessential Mr Intel as king, Chipzilla is betting on being more Intel than ever. But it's also getting someone with an outside perspective and the moral authority within the company to use that to drive change. It will be Gelsinger's company, and many of the cultural and structural impediments to executing his strategy just won't apply. He'll execute all right – but what?
Intel is flying in coffin corner, that feared flight regime where an aircraft can't go faster without exceeding maximum safe speed, and can't go slower without stalling and falling out of the sky. Intel can't go on failing to make chips in the numbers or process the market needs, and it can't go on investing in things that just don't work. It needs to point in a new direction, but that has never been what makes Intel Intel. Yet it can't stop being Intel. How to square that circle? Nobody knows.
I said earlier that Pat Gelsinger was transparently decent, a point that many godless Brits conflate with his Christianity, something he lives with an openness unusual even for an American. These things are connected but not contingent, he's not the sort who needs to get his morals from a book. But what he does have is confidence in faith, and faith is the ability to move ahead confidently in the absence of knowledge.
He won't know what's going to work, just that he's got it right in the past, that he knows the technology, the market, and the people. That's a potent mix to bring to a hard task – and however it works out, it will be a unique and fascinating story in the unfolding.
In an industry so uncertain of itself and its role in the world it has created, a man with faith in himself, in possibilities and the things he's been given to work with, will be a man to watch.
And if it doesn't work out, there's always that career in technical writing. ®