'Zeno Arrow' said:
Indeed. RISC architecture is gonna change everything.
technically, risc changed the world of computing eons ago. it is just until now there were way too many blinders for users in the personal computing sector (particularly, on the pc desktop) to realize that. the two deadly sins of the x86 related to that are:
* it tries to emulate a risc-predating archtiecture, with all the drawbacks coming from that - unnecessary boilerplate decoders/frontends, effectively serving as roadblocks for compilers in doing their job, etc. ia64 was a step in the right direction, alas it still carries x86s bad-gene heritage.
* power efficiency is a relatively new trend for intel - they realised it mattered only circa p4 - the cadillac of cpus - 5 gallons per mile. yyyyhaaaa, cowboy!
the p4 market fiasco was a wakeup slap in the face for intel, though, forcing them to realise they needed to mend their ways. so now they are trying to address their first deadly sin - them ignoring power efficiency in the personal computing sector. their second deadly sin - their assumption that people who care more about running decades-old code, rather than having their todays software run optimally (basically, people like the thread starter), are the majority in the personal computing domain (an assumtion which is plain wrong) still pulls them down, though.
the funny thing is that, even disregarding any oss communities where software portability/migratability has never been much of an issue, purely consumer-oriented personal-computing companies like Be inc and Apple have been proving time and again that having a well-maintained, modern code-base for their products is _way_ more important for the end user satisfaction, than keeping binary compatibilty ad infinitum.
of course, all of the above intels sins - collectively-called "wintel deadlock" - were actually a golden egg for the chip maker, and they actively sought preserving that status quo for a long time (why not - it was bringing them profit while allowing them to stand still). unfortunately for them, though, times change. people today expect from their hardware things that are fundamentally-incompatible with wintels golden-egg philosophy. thus, enter the current situation where intel are desperately playing a catch-up game with the rest of the cpu world in the most "personal" of the computing segments - handhelds/pocketables; do not be fooled about the size of this market - it includes _way_ more than geek gadgets - these are also (smart)phones, pocket consoles, PDAs - all multi-billion dollar markets. unfortunately, intel still have not renounced from their second deadly sin, and that may cost them dearly.