I can't say for sure; no insider information from here in the Valley.

I could reasonably theorize that it's simply the vast disparity in size and muscle. It's not just raw power, but also synergistic effects (gack, I'm slipping into bizspeak; please kill me if I try to show you a PowerPoint presentation, OK?). Silicon Valley's changed dramatically over the last few decades, from a place built by and for entrepreneurs, into a place controlled by networks of interlocking boards and alliances and partnerships.

It looks, in fact, like an adaption of the Japanese zaibatsu system. The major change seems to be that instead of directly owning a bank, the the gigacorp at the top owns a venture capital fund. No doubt US banking laws are responsible for the adaptation. Intel has one of the biggest VC funds around, and that's part of why it's even bigger than it seems.

The other striking adaptation is intimately related to the VC funds: The institutionalization of startups. Very few are actually started in a garage by people rebelling against their employers, the way Woz did when he left HP to found Apple. Nowadays, people toil inside a big corp for a few years to bulid up contacts and a reputation, then leave to set up a startup pretty much with the blessing of their former employer, who often connects them up with its VC fund. Even weirder, if you still hold any idealistic freemarket notions, is the way that these corporate-entrepreneurs are passed along the VC network, from fund to fund, even across the border between supposed competitors. The VC funds don't always do it for money, either; in the early stages, each fund might instead take a few points for making introductions to other startups and businesses in its portfolio which might create "synergies" between the startups. It's the old-boy network on a sound commission basis.

The system has considerable advantages for the corporate sponsors, because it compartmentalizes the risks and uncertainties of R&D. When you take those risks in-house, it directly hurts your bottom line when it doesn't work. But if you farm out the risks to startups and assign any losses to a VC fund, the corp's balance sheet looks healthier and Wall St. is happier.

To get back to your point, I think it's simply the case that Intel is both powerful and secure in its web of business relationships, and has considerable power to not only kick AMD in the marketplace, but also to constrain its ability to form alliances and partnerships that might give it added strength. Any company tempted to help AMD has to consider how Intel might punish it some way, if nothing else by passively starving it of support and capital.

That's my theory, based on observation, sometimes from inside the kind of startup I described.