Will this be basically a bit like when Commodore brought 6502 CPU maker MOS - in the olden days?
Mines the one with a badly researched book about the history of computing in the pocket.
With Intel and Advanced Micro Devices both cranking the clocks on their respective Xeon and Opteron server processors this week and Dell admitting that it's on the prowl for acquisitions, tongues have started to wag about the possibility of PC and server maker Dell snapping up AMD. Barron's, the supposedly more thoughtful part …
(which probably means it will happen)
Not only would it pretty much force Dell to stop using Intel (even if Intel wanted to keep doing business with AMDell - which, because money is involved, it might do anyway) but I would imagine that many people would stop using AMD, not out of spite, but because they could not be sure of the long-term support that they might receive for their chips in the future.
I would imagine that Dell would be much better off, as stated in the article, letting Intel and AMD keep themselves busy providing chips with better performance and cheaper price.
Another thought would be if Dell wanted to invest in new FAB buildings - then rent them out to Intel, AMD or whoever wants to make CPU, RAM or GPUs or even all-in-one chips.
"The minute Dell bought AMD, it would inherently be worth less money because Dell would only want AMD to make its own chips for servers, storage, and PCs and leverage these as an advantage it held over its competitors."
Is Dell's market share similar to AMD's? That is, *could* Dell consume all of AMD's output? If not, they'd have to sell the remainder to somebody. Either way, AMD's offerings would still have to remain competitive or else AMDell's offerings would not be.
Also, the value of the AMD portion of AMDell only matters if Dell need to re-sell later on, at which point the AMD portion is worth whatever it's products are then worth in the open market, which again depends only on how good they are.
AMDell's advantage would be that they could design whole systems. Merely producing a great CPU is no longer enough. The bottlenecks are all between the processing and the data, so optimal performance will be achieved by custom design at that (wider) scope. This acquisition makes more sense now than it would have done 10 or 20 years ago.
I'm still not sure if it makes sense now, mind. Unfortunately, it's not my money to spend.
This could too easily play-out like Sun Microsystems when it switched from Motorola 68k to SPARC. The example is instructive because Sun ended up with a vertical marketing strategy that saw it move from the “new apple” (not so ridiculous when you consider NeWS early promise) to a builder of bigger and bigger servers.
Dell would only buy AMD, if it was a bargain.. and it is not that yet.
Intel-AMD cross-licensing agreement says that if AMD goes under, gets acquired, or goes private, all AMD patents go to Intel and buyer of AMD is left with no x86 license. Can't produce chips.
Especially under section 6. that says EFFECTIVE DATE, TERM AND TERMINATION FOR CAUSE
My take on why ORACLE bought Sun was because Sun's chips had RAS (reliability, availability, serviceability) features, and ORACLE wanted the ability to go absolutely head-to-head with IBM in the large mainframe database business.
Nehalem-EX, which brought RAS to the x86 platform, probably had Larry Ellison kicking himself for wasting money.
So if DELL doesn't need AMD for anything, the same applies to ORACLE. Now, IBM and Apple, mind you, could probably put AMD to some good use. But that isn't even rumored, and there are good reasons to think that they aren't really in the market. Besides, the important part of AMD has already been sold off to Dubai.
They spun off all those assets to Lenovo etc., ages ago. They want to be a service agregator in that market. So far they've done well with that. Why would they reverse course now? Yes, they still compete in the big iron, but that's a different market segment.
I was trying to think of someone who can buy AMD... Dell? no, Oracle? no, HP? no, I was trying to remember who I thought about before until I read an article about CISCo and Intel recently. Pow! yeah, it can be CISCO... CISCO USE INTEL CHIPS IN ROUTERS AND SWITCHES , WHY EXPENSIVE INTEL CHIPS?? WHILE AMD CHIPS WILL DO..
A Private Equity takeover would exhibit none of the difficulties that you point out in your article.
Additionally if AMD was not absorbed in the fashion that say ATI was but rather became a division say the way that Carrier is a division of United Technologies then the licensing aggreements would probably hold within the division.
Rather than Dell, a perfectly good buyer of AMD would be ARM Holdings. ARM seems to desire a PC presence but their RISC 32 bit cpu's really can't compete with 64 bit x86 which will soon evolve into 128 bit with the release of AMD Bulldozer.
With AMD they would then have the entire CPU market covered from mobiles to servers not to mention Supercomputers with x86 AND HPC GPU computing using ATI designs. It satisfies the observation that white box makers would run AMD with any merger with a competitor while at the same time keeping the company in the business of CPU's, AMD's wheelhouse. It would also strengthen both ARM Holdings and AMD against Intel. Who I am sure would run off screaming into night.
There is also the little commented fact that Intel will begin dropping the PCI bus in about 4 years. The agreement with AMD sees Intel support PCI for only 5 years. That suggests that Intel is developing a proprietary bus to replace PCI and licensing to AMD may be a huge profit source.
The question is though could ARM Holdings raise the cash?