IBM to bleed Intel ?

Before delving into the core of our subject, i shall begin with a little foreword. What follows is a rumour. It is in no case a change in regards to our decision not to publish rumours on Apple and its products. It does concern the Cupertino based firm only from afar, and cannot have any influence on the course of things either in the short, mid or long run. As for all rumours, this one could be wrong. But its content gives it enough interest so that we've decided to publish it.
IBM is nowadays very active in the ground of high research on processors. The firm has managed -thanks to its numerous discoveries- to secure a promising future. Its most achieved and beautiful patent is the SOI, which allows to quite easily gain 20% on processor frequency without any other major change. It is the SOI that has partially helped the G4 reach better frequencies and gain power. AMD and, more recently, Nvidia have obtained a license to use the SOI in their own products. Intel had chosen other paths which also seemed promising. Eventually they revealed to be less promising than expected. Since a few months ago, the CPU giant has ceased its frenzy race for faster extreme frequencies. True, the passing to 90 nm spurs temporarily the rise of frequencies, but in the long run and when passing to 65 nm, it will be in a bind.
Let's get back to the core of my discourse: Intel, just as AMD and Nvidia did, would have asked for an SOI license. Whereas the two others pay close to nothing, IBM would have asked Intel to pay US$ 500 Millions plus US$ 45 per manufactured processor. Intel has thus the choice between losing many months in research and development of another solution OR finding itself selling its processors at higher prices than... AMD.
We let you imagine the implications if this rumour revealed to be true.