There is a fable being passed around to the effect that "iPad is killing off the PC market". It is untrue. There is no monolithic "PC market" for starters. Rather there are several very different segments where x86 CPUs are consumed. In the vast majority of those segments, iPad isn't even a participant much less a threat.
The most important segments are: consumer, enterprise and server. Each of those segments are comprised of sub-segments. iPad is obviously not a factor in the server segment. Similarly, few if any enterprises have replaced laptops and PCs with iPads. That means that iPad cannibalization of x86 marketshare is isolated to the consumer segment.
But even in the consumer segment, tablets like the iPad are of limited usefulness for anything but the most mundane tasks. They aren't any good for high end gaming, nor for graphics intensive applications such as digital photography and movie making and editing.
iPads are suitable replacements only for low end, de-contented laptops and PCs, therefore. In the lowest end of the consumer segment, there is no doubt whatsoever that iPad is wreaking havoc. To the extent an x86 chip maker's revenues depend on this segment, it is subject to massive disruption of revenues that is (a) wholly unrelated to general economic sluggishness, and (b) will only be accelerated and exacerbated should the economic environment improve. How's that? Well, iPads are the preferred solution now for low end consumer needs. That trend exists in both good times and bad, and the growth of iPad at the expense of PCs in this segment is an irreversible trend. Good times merely make consumers more likely to buy what they really want, which is an iPad. Economic improvement, therefore, won't help x86 CPU sales in this segment.
So who is exposed in this segment? Both Intel and AMD play in the low end consumer segment. But only AMD depends on this segment for the vast majority of all of its revenues (x86 and CPU). This means that iPad stings Intel but lays bone-crushing blows on AMD. As you move out of the low end of the consumer segment, Intel is utterly dominant. Ivy Bridge is so much better than anything AMD sells presently or has on its roadmap, that Intel has been able to gobble up market share at the top of the consumer segment every single quarter for the last three quarters. That won't change either, because AMD doesn't have the resources to bridge a 2-generation CPU technology deficit.
AMD has never been well-represented in the enterprise PC segment. Intel has always and will always dominate this segment. Note that this is an iPad free zone. No worries about ARM cutting into market share here. At least not for a couple of years. The same technical inferiority of AMD's CPU products that is killing it in the high end consumer segment prevents it from having any chance fo success in the enterprise segment. Economic improvement will stimulate sales in this segment which in the x86 context benefits Intel almost to the exclusion of AMD. No help there.
That leaves servers. Each quarter for the last several quarters, despite major much-hyped Opetron launches, Intel has been taking one-to-two basis points of market share from AMD inthe server segment. We don't have the industry marketshare data for Q4 for the server segment yet, but we do know one important thing. In Q4 Intel significantly grew its server revenues; +6% according to theier report yesterday. That's the largest single quarter growth in that segment in a year, and if in past quarters where Intel's server revenue growth has been more modest, it has taken away server market share from AMD, I think it's inescapable that it has done so again - probably to a larger extent - in Q4. Like the enterprise segment, a rebound will merely accelerate the trend we see away from AMD and to Intel, so again, rising tide won't lift AMD's boat here either.
So what's the net-net? AMD is being APPLEHOLED from below and IVY BRIDGE HOLED from above. Economic recovery will do nothing but accelerate the trends already devastating AMD during these lean times. There is absolutely no hope of optimism for AMD in any of its core revenue generating businesses. AMD's survival depends on AMD, a chronic failure, magically avoiding all of its usual screw ups and miraculously delivering a compelling product in a segment where it has never competed before, in competition with much larger, more successful, better managed, and much more reliably executing companies like Samsung and Qualcomm.
In other words, AMD is up self-cornhole creek without an AMD-logoed buttplug.
AMD is paying double for its CPUs since the revised agreement with GF, and now it's losing market share, too?
Back up the truck, I'll get the shovel, and we'll scrape AMD off the double yellow line.
If AMD is also going to source more at TSMC, this gives TSMC the opportunity to mark up for their trouble.
TSMC does not have to give AMD any leeway to negotiate wafer pricing.
Stupid AMD fanboys.