Intel has failed in most of the markets other than their core CPU business. Examples: Remember the ethernet switch project - capilano which was a huge disaster. Their GPUs are way inferior to nVidia/AMD, they haven't made any headway in the ultra-mobile space.
Goes to prove that having deep pockets is not the sufficient condition to entering a market and being succesful in it.
It might be Intel's complacency which made them think they the way ruled the PC world they could reign the mobile world too. The other players have moved much faster and offered better solutions that the mobile world has almost forgotten Intel. Similar fate has been meted out to Windows. So in the mobile world WINTEL has lost the race.
my feeling is intel has reaped 2 much from cpu monopoly.
glad to see it's feeding some junk talent with it's extra $$s. this is a way they feed back the society as well. ie. those folks don't need to get government aid, intel ll pay them.
Yes, but the question is, what is the right decision?
I thought it was interesting to see that what would normally make sense, i.e. integrating the CPU and RF parts into a single chip, for handheld devices, is not necessarily your friend. This is too volatile (fickle) a market. Things change too fast. And the handset/tablet vendors apparently don't like to be tied to one chip vendor only.
Seems a little like the manufacturing problem. Apple chooses to stay in China why? Because they can provide a more manual manufacturing process, which is faster to restructure, for the same cost as highly automated plants in the US. Or maybe even less. And being able to change the manufacturing process is essential to keep up with the fast-changing products.
So, counterintuitively, best not to obsess too much about integration to the nth degree? Allow your solutions to work more flexibly, with other chips? Give the vendor an easier time of creating upscale and cheaper products, mixing and matching components as needed?
Almost seems like Intel would be better off creating a whole launder list of more single-purpose building blocks.
It is really hard to call it a struggling giant, based on http://www.eetimes.com/electronics-news/4369843/Chip-rankings--Intel-had-highest-share-in-over-10-years.
I guess ARM threat is a bit exaggerated.
Blog Doing Math in FPGAs Tom Burke 24 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...