Apple needs more leverage!!!
If Apple's market share were 20%, do you think that AIM would have failed to live up to our hopes? I don't think so.
Apple wasn't buying enough CPU's from either IBM or Motorola to dictate their strategic directions, respectively.
BTW, Apple moving more of its products to Motorola's G4 increases volume and hence Apple's influence over Motorola's strategic direction. This is something of a speculative strategy in that it lessens Apple's influence and relationship with IBM. Fortunately, Apple may have the option of purchasing the intellectual fruits of its alliance with Motorola and outsourcing to IBM or another manufacturer.
In fact, compared to their embedded markets, I believe Apple is a small fish. Does anybody know what kind of overall CPU volume that IBM and Motorola each crank out per year in PowerPC chips?
As I've said before, the embedded market, and to a lesser extent IBM's server market, prefer low power consumption over raw computational speed. While this point is arguable as to the degree that this is so, it clearly is a fact.
We can bitch about Motorola and IBM failing to produce PowerPC chips that enable Apple to indisputably kick butt in the PC world. But, what Apple needs is both a top-notch semiconductor supplier and enough market power to motivate that supplier to devote the necessary resources to delivering the devices that Apple needs to kick butt.
Now, if Apple waived a magic wand and instantly had 20% market share, the semiconductor supplier issue would practically tend to itself. Well, that is not going to happen!
In lieu of instant market share, Apple has two options to increase its volume high enough to COMMAND the devotion of a top-notch semiconductor supplier.
First, Apple could guarantee the purchase of a minimum number of CPU's, regardless of whether Apple resells them in PC's or not. This speculative strategy probably isn't in the realm of practicality.
I say 'probably' because Apple might be able to (may be contractually forbidden) resell their excess CPU's to embedded customers, such as Cisco, below cost. Even so, this speculative strategy is extremely risky. I probably wouldn't do it unless the numbers increased my confidence as well as if there could be a short-term radical performance improvement as a consequence of this volume committment.
The second approach is less speculative. Apple would seek partners of like-needs. This strategic alliance would create the demand for the dream G5 or whatever that Apple has been unable to do so alone.
But who would Apple partner with such that they as a collective could command sufficient volume-demand to motivate a semiconductor supplier to achieve the desired results? SGI? Sun? IBM?
I don't know what these and other companies are doing and thinking about with regard to CPU's. If you know something about them, you might contribute to this discussion by shedding some light such that we might conceptualize a strategic alliance formed for the creation of the dream G5.
Well, enough with this post.
Eirik