Intel Corp (INTC)

[Unkwn]

he lack of an integrated modem had nothing to do with contra revenue. No one else has an integrated modem

Well, depends on your definition of a modem. Call it communication module or whatever you like, but it is a fact that Qualcomm and the likes have a much higher degree of integration, which includes the digital part of the modem. The analog frontend is always (?) external, I know that. I don't have an explanation for this, since it should be only digital logic, but I guess I am (we are) missing something here, otherwise it wouldn't be as hard for Intel to integrate this part (and opt for TSMC to fab SOFIA instead, with the plan to go back to Intel's fabs once technically possible).

In any case, this issue has nothing to do with FPGAs, not even close.

As I said earlier BayTrail was optimized for higher end Windows tablets and not for lower end Android devices. I believe that the first generation Baytrail chips only worked with higher end, and expensive memory chips and not with cheaper memory devices, and so Intel offered a subsidy to offset the cost of more expensive memory devices.

I think you're believing too many of those excuses Intel names officially. If that would be true, why did Merrifield and Moorefield fail so badly, since those were targeting the smartphone market right from the start? It seems the integration and die area field is where Intel fails, not some "designed for the wrong platform" BS. Again, in every case, not something that can be fixed with FPGAs.

But it may not be able to anticipate the latest trend in peripherals and system configurations. The on chip FPGA gives much greater flexibility.

Yeah, more flexibility by sacrificing about 1000 times the die area and power. For mobile? Doubt it. For some highly sophistic server extensions? Maybe (but off die, with PCIe should do the same already).

don't think the Altera talks are about not being able to compete without buying the company, I believe that Intel has finally realized that the most profitable way to take advantage of the process lead that Intel enjoys is to outright purchase high margin chip companies and reap 100% of the benefit that companies gain from access to Intel's Foundry.

Well, Intel can't buy every company in the semiconductor business that have a good margin. These companies are usually also one thing: expensive. Intel needs its money more than ever before to pay for new fabs and Intel doesn't particularly have a good track record about integrating other companies. They seem to insist on their way of managing a company, which I fear is the wrong way to run Altera. Altera is a perfectly run business, so why change it?

I really hope Intel is only planning to buy a significant share of Altera to intensify their technical cooperation. That would make much more sense to me.