I see a lot of speculation with regard to a potential acquisition of AMD by the likes of Qualcomm (QCOM) or Samsung (OTC:SSNLF). These kinds of rumors aren't new and pop up every so often, usually when the stock gets too low. The aforementioned players are usually brought up as potential acquirers, but there is very little truth to these rumors. Quite frankly, an AMD buyout at this stage of the game - especially as the growth segments (smartphone, tablet) of compute can be services with an ARM (ARMH) core/architecture license - makes almost no sense from a technological perspective, and certainly not from a financial/business perspective. Here's why.
Potential Acquirers Can Poach The Talent
X86 compatibility was a big deal for computing before the smartphone revolution. Without it, there was no real mainstream consumer computing segment that could be serviced with anything but a processor from Intel (INTC). Nowadays, nearly all platforms (besides the PC) can run on MIPS, ARM, or X86 without a hitch. To do a tablet chip, a company can simply license the ARM instruction set and build its own core + system-on-chip ("SoC") (which is what companies like Qualcomm do).
So, as AMD has struggled, companies like Samsung have swept in and poached a lot of the AMD talent to presumably help design their next generation ARM-compatible processor cores. In particular, if you look at the LinkedIn (LNKD) profile of the chief architect of AMD's "Jaguar" core, you will see that he is now a proud employee of Samsung's Austin design team:
Intel CEO Paul Otellini noted at the recent Bernstein conference that Samsung essentially hired the AMD CPU team in Austin for its micro-server chip development, which seems to jive with my poking around on LinkedIn. If any company wants the AMD talent, then they can simply poach it, as these larger companies offer security and likely more generous compensation than AMD can afford.
AMD's IP Isn't Worth Much
As I said, while X86 based IP might be nice to have, there would likely be significant complications with regard to the cross licensing deal with Intel? For the new growth segments, ARM or MIPS works fine, and in the PC space, competing against Intel is just a bad idea. AMD has some good graphics IP for PCs and potentially HPC, but for the most part, any client graphics needs could be serviced by licensing IP from ARM or Imagination...or simply developing it in-house by poaching the relevant AMD engineers.
Quite frankly, there's not a lot of money in AMD's IP given the state of the processor world today. If this were pre-iPhone, or if the mobile world had originally been X86 based, AMD would probably be doing so well as a second source that nobody would even be talking about an AMD buyout in the first place! But with ARM licenses available, and even ARM core designs out there to be used, just what value would an AMD acquisition really add to any of the giants in this space?
AMD Has A Lot Of Debt
Another problem with a potential AMD acquisition is that the acquirer would be, in addition to picking up a business that isn't generating positive free cash flow, taking on a tad over $2B in debt. So even at today's "cheap" market capitalization of $1.9B, the acquirer would be liable for an additional $1.04B in net debt (this doesn't even include any potential tax implications from repatriating any overseas cash). Sure, that's not a lot of money to a company like Qualcomm, Samsung, or Intel, but these companies didn't get to where they are today by making poor financial decisions.
This wouldn't be a deal breaker if there were significant IP value to be had here, but there isn't - the debt just makes it that much less attractive.
Don't buy AMD speculating on a buyout. AMD needs to either fight through its problems on its own with its new strategy, or it will eventually go under and its patents sold for pennies on the dollar to service debt. I'm betting that AMD will probably be able to reinvent itself and survive, especially as it tries to stay more and more out of Intel's way, but it is a speculative position.
Additional disclosure: I am short ARMH