Nice post.
From my point of view your after table says it all:
After:
1. Dell (INTEL)
2. HP (INTEL+AMD)
3. Lenovo (INTEL+AMD)
4. Acer (INTEL+AMD)
5. Fujitsu-Siemens (INTEL+AMD)
When I started investing in AMD there was only a glimmer of hope that AMD would be able to come up with a competitive processor design and only in the desktop/single server area. In the processing/manufacturing/distribution areas AMD was hopelessly outclassed, and in the business/marketing areas AMD was operating in some alternate universe.
Today, AMD has/or soon will have competitive or better products than INTC across the entire processor spectrum. In the manufacturing area 90nm is helping, but fab36 will be needed to really make AMD competitive. Distribution still needs work, and despite the seeming improvements in processing lately I would say the jury is still out (too much very bad processing history to get excited here, but IBM relationship has seemingly done wonders). Business/marketing is the area that still concerns me the most, though as you point out, things seem to be getting better and better.
I think you're right in regards to AMD giving up on DELL. The thing I found most interesting about that whole farce was the apparent angst DELL went through getting to the decision to stay with INTC only. Now if someone with as sweet a deal as DELL has with INTC could even contemplate adding AMD you have to wonder about the rest of the guys out there. As I said in my previous post the incentives to cheat seem to be overwhelming the risks. In short, INTC's ability to discipline truant troops seems to be a lot less obvious than it was, say 6 months ago.
As far as Lenovo going with AMD to distinguish itself from DELL. Frankly, I expected this course of action from HP. Though it does now appear that HP is now warming up to AMD, by now I had expected the relationship to be torrid. In the end I think this was Carly's biggest mistake.
IBM seems even worse than HP. Their furtive steps have belied their outward messages of support. Further decision making at IBM seems to be one of the most convoluted and arduous processes imaginable.
I'm not really surprised at the convulsions big companies go through in making decisions, but it's no wonder the little guys can run circles around them.
Again, I think your right about INTC eyeing IBM's Power business, which should be driving IBM into AMD’s open arms. But as I've said, I expect this corner of the market to come under increasing pressure from below over time. The Opterons seem to be evolving into a platform that can compete in everything from the single server, to the enterprise, to the super computer markets.
Finally, your estimates of market share gains seem reasonable. Realistically, at present, there is no way AMD can gain more than 1 to 1.5% of market per quarter. But even at that pace it would do wonders for AMD's largely fixed cost business.
