David Kanter (dkanter.delete@this.realworldtech.com) on August 1, 2018 10:44 pm wrote:

> > This may only be the ravings of a daft fool over here in
> > Ireland, but the fact that this story about Knights
> > Landing was the major breaking news about Intel's innovation
> > in 2014, is the whole problem. The whole problem
> > is that it's 'B' team even ended up on a project like this.
>
> So you think they should have just had their A teams (Oregon and
> Israel) stop work on the big cores that make all the money?
>

Xeon Phi project was too small for either Intel's A team, B team, C team, or any team they could have put on that field. Period. The fact that no one, either on analyst side or internal management could reach that conclusion is a problem. It's not that building datacenters is outside of the core competency of a company such as Intel. It's the 'size' of opportunity that Intel as a corporation should have been looking at. Not stuff, that is better left to startup funding. Insert any other 'large' project in there, instead of datacenters, but that's one area that requires a large amount of capital investment and management. That's where Intel's core competency is. But it doesn't have to be related only to building of silicon foundries. And shareholders and analysts like you, ought to demand more of the corporation that just that. That's my argument, disagree with it if you wish. It's still a legitimate argument.

There was an implication made above, that a 'B' Intel team wasn't very good, or was incompetent in some way. And if that is a perception about Intel from the outside, then yeah, I'd say that is a perception that is worth changing from within. I.e. The 'B' team at Intel ought to aim to be the equivalent of 'A' teams anywhere else. And 'B' teams at Intel certainly ought to have something more ambitious to aim for, than a project like Xeon Phi.

2014, when you original published this article on Realworldtech was a time period during which Intel could have focused it's considerable resources and talent a several other worthwhile engineering projects. And it didn't. It got stuck with this. Maybe, the reality that you (i.e. David Kanter), can't look at Intel as anything else other, than the kind of company that it's always been (in your mind)? And is that part of the whole broader problem? Intel being a company which is now surrounded by people who simply repeat to it, stuff that it wants to hear? Do you think there's some argument to be made for that, or not?

And here's a question I'd ask while we're at it, does Intel in 2018 not start to look like a company, not only from a previous decade, but from a previous century? Do you not agree, that is the kind of challenging question that you ought to ask? As an analyst with a genuine and sympathetic concern for the corporation, that you know so well?

>
> Intel is a semiconductor manufacturing company. They are in the business of manufacturing
> and selling processors/chipsets/networking/storage to data centers. Sometimes whole
> servers. They don't operate data centers, so why does this make sense?
>

Or is that the traditional cultural bias that you bring to the table with you, which doesn't make as much sense any longer? And is that doing any good for the engineers and the industry that you like so much, and have been fascinated by for so many years now?

> David