As far as Monday morning news goes, the email from Pat Gelsinger telling of his departure from Intel was something of a shock.

"After almost 30 years with Intel, I've made a monumental change. On Monday, September 14th, it was announced that I would be retiring from Intel and joining EMC," he wrote.

In fewer than 30 words Gelsinger ended a relationship that had lasted for 30 years. So why did he do it, and what does it mean for Intel?

Gelsinger started as a quality and assurance technician at Intel in 1979. His relationship with the company might be the longest of any in his life, and walking away must have been a huge decision.

The first question is, did he walk or was he pushed? I'm inclined to believe he walked. There's no obvious political reason for him to be forced out, nor any evidence of any such move.

Gelsinger was most obviously happy when he was in deep geek mode. As chief technology officer, and with his post in the research division, he was best suited to the job.

When he moved over to enterprise computing he was noticeably less comfortable. Sure, he could talk the talk but Gelsinger is nothing if not interested by pure technology, and the move to pushing server systems seemed to irk him.

For Gelsinger comes across as a geek, in the purest sense of the word. He was never happier than explaining new technology and speculating on how it would change lives.

Probably a key factor in Gelsinger's decision to move was the realisation that he wasn't going to get the top job, not because he didn't deserve it but b ecause Intel isn't a technology-led company any more.

Intel was founded on Gordon Moore and Robert Noyce, hardcore techies who saw the opportunities of a life in silicon. Former chief executives Andrew Grove and Craig Barrett, meanwhile, carried on the tradition of having a PhD in physical sciences at the head of the company.

When the present incumbent Paul Otellini was appointed as chief executive, it was a major shift for Intel. Otellini wasn't a technologist, he was a money man. For Otellini the business was the thing, not the technology.

After the embarrassing Itanium episode, Intel felt that it needed to step away from focusing on pure technology, and Otellini has done a good job keeping Intel on a steady road to growth.

In the old days, Gelsinger would have been in line for the top spot. Now it's not so cut and dried, and if I was a betting man I'd put money on Sean Maloney taking over from Otellini. Maloney, while strong on technology, comes from sales and marketing and that's where Intel will find its next leader.

So Gelsinger goes to EMC. He should be on course to run the company in a few years and, to be honest, EMC looks a lot more technologically interesting than the chip business at the moment.

Storage and virtualisation are where it's at nowadays, whereas the chip business is largely a manufacturing concern. Compare the suited and booted crowd at SemiCon with the open atmosphere of VMworld, and you can see where Gelsinger might be happier.

Intel will be a poorer place without Pat walking the halls, and next week's Intel Developer Forum in San Francisco will be terribly dull without him.