I seriously doubt Intel would succeed with this acquisition, unless the goal is to kill it. Replacing Atom with an ARM based architecture makes no sense whatsoever. That was the whole reason Intel sold Xscale to Marvell, right? The Altera acquisition I understand but QCOMM would be a major folly under current Intel centric management. My opinion.
Moving forward my advice to QCOMM is to go on an acquisition spree like no other. Get into the server business, networking, everything else that mobile enables. My opinion.
Qualcomm is in trouble, but so is Intel.
First, Qualcomm can not acquire a server maker. There are two real server instruction sets, x86-64 and POWER. I guess SPARC too, to a lesser extent. x86-64 would be really difficult, since VIA/AMD/Intel have it, and VIA isn't even close to making a server chip. AMD would be highly problematic, because they'd have to re-negotiate with Intel to get the x86 license, since the original agreement would be voided. Even then, AMD has a horrible current architecture that simply doesn't work, and it isn't clear how well Zen will perform. Clearly it will be better than Bulldozer though.
QCOM certainly can't buy Oracle, or IBM, so they don't have a clear way to purchase their way into the server market. They are trying to build ARM server chips, but, that's a long way off, and they have no experience with it. On top of that, server customers are very, very cautious about moving to another architecture, so this is a very long term play, and is unlikely to be successful in any case.
OpenPOWER is another possibility, but IBM makes the processors, and everyone else makes other parts. Except for a Chinese company IBM licensed the instruction set to. It's unlikely they'd license it to QCOM, since POWER8 is extremely capable, and far beyond anything QCOM has ever dealt with.
For everything else, other companies are also aggressively going after those segments, including Intel. Competing against Intel is extremely difficult, not because their products are very good, but because they need to keep their fabs busy, and will price very aggressively to make that happen. So, good luck there.
For Intel, it might make sense because they need to keep their fabs busy, and QCOM could. Except, we already see how long it has taken to move their Infineon IP to their own fabs. Plus, Intel's 14nm is a dog, with current high-end processors overclocking at least 300 MHz lower than the 22nm node. Also, with Apple making 92% of the phone profits, and Samsung making 15%, clearly the companies making the money make their own CPUs.
Sure, QCOM has importance in modems, but that's getting to be more of a commodity for 4G. Also, Intel is in that market as well, making the overlap problematic. As mentioned, moving it to Intel's fabs, keeping mind the glacial pace that company operates at, would be slow as to make it unimportant. Most likely, Intel would kill it, since they have their own technology on their own fabs.
I don't see it. Two sick companies don't make a strong one. I'd stay away from both. TSMC makes much more sense if you're looking at companies around that market cap.
