Why Intel Needs Smartphones More Than They Need Intel 134
An anonymous reader writes with this excerpt from ZDNet:
"The launch of the Orange San Diego, the first handset using an Intel Atom processor, marks a big milestone for the chipmaker: it's finally in the smartphone market. But does the market need Intel at all? ... Intel's scale and the reach of its other divisions gives [Mike] Bell's smartphone unit a boost; for example, it can reuse code optimizations for Atom done by the desktop team. ... Even so, the smartphone team has got a tough job on its hands — but it's one Intel has to tackle, according to Carolina Milanesi, mobile analyst at Gartner. 'This is certainly an attack strategy for Intel. The smartphone market is so large now that they need a piece of the pie,' she said. But will consumers care whether their handset runs on an Intel chip? Bell conceded that aside from the tech-savvy, most people probably don't know which chip is inside their phone. It's likely, given the lack of advertising on this, that most probably don't care — making Intel's job even harder."
Games? (Score:5, Interesting)
There is games, though; some Android game engines are written in part in native code for the speed boost, and I can't imagine that an Intel phone will shine when forced to emulate an ARM CPU on the fly for those occasions.
And for most applications, the CPU really does not matter. They'll run nicely on anything able to host the Dalvik VM. At best, an Intel phone will be no different than a ARM one, and at worst it will just add an extra bit of frustration.
always protect the low end (Score:5, Interesting)
performance wise ARM is crap compared to Intel. Just like Intel was crap compared to SPARC and all the other architectures they killed off in the last 30 years
one of the most important rules of business is to protect the low end of your market. if you don't then a competitor will establish a lower margin business and move up to take your high end. Just like Intel did.
even apple knows this and has products just good enough to keep low end competitors at bay
Re:Recompile (Score:4, Interesting)
Easier said than done. You have to test and support both versions, and recompiling C/C++ cross platform is not always straightforward. Given the already significant fragmentation in Android, I wonder when/if many places will get around to it. (The answer is when Intel gets enough market share)
Re:Does the Market need Intel at All? (Score:2, Interesting)
The ARM platform is not one device. Since ARM is an IP house only there are many ARM platforms out there: Single core, multi core, we'll see 64-bit ARM soon. Then there are all the integrated peripherals around the CPU core: DSP, video CODECs, camera interface, etc. What will be interesting is that if you are bought into the Intel ecosystem then you'll only have Intel-based solutions, there will be no competition within that domain. ARM is amazingly exciting with all the offerings you can get. The designer (and thus consumer) has a lot of choices in the ARM marketspace for trading computing power for battery life.
And to the person that commented that ARM performance sucks - Well the ARMs NEON architecture is pretty sucky - but that's really not what the applications are necessarily depending on. My iPhone 4s is quite nimble at serving up webpages, playing games, streaming movies off my NAS drive. Intel's offerings will probably be just fine technically (OK, probably lacking in graphics area, but...) - It will come down to cost and features. IMHO it was stupid for Intel to have ditched its Strong-ARM holdings. Absolutely stupid. There is no advantage by having only one type of CPU architecture model - that's why we have high level languages and compilers.
Comment removed (Score:4, Interesting)