Replies

Remember when RISC was supposed to be the future of high speed computing?

Almost every cell phone in the world runs RISC, and the smartphones are getting to be faster than the notebooks of only a few years ago while using a tiny fraction of their power. x86 dominated the "high-speed" portion for a while, but these days people are more interested in the highest performance for a specific power draw. People want it for their battery-powered devices, businesses want it because power is expensive. RISC absolutely dominates in that calculation.

People want it for their battery-powered devices, businesses want it because power is expensive. RISC absolutely dominates in that calculation.

I am aware of ARM and their dominance in the Smartphone market. Interestingly, Microsoft, who used to back Windows NT for a variety of processors, including RISC processors, backed out of that (MIPS, Alpha) and of course Apple dumped the PowerPC for Intel. This almost certainly has more to do with market forces (IBM was NOT going to let Apple get enough chips in a timely manner, Intel welcomed the new customer, and Apple has AMD as an alternative if it is ever necessary. Plus, the $$ savings at the hardware level by having standardization on x86 / Intel x64).

So what we have is CISC continuing to dominate the desktop/workstation/notebook market (including Linux boxes), splitting on netbooks (Intel Atom has a healthy share), and competely losing out on the smartphone, iPad, smart gizmos end. I don't know if the Itanium is truly RISC, or if anyone really runs it. I assume IBM's Cell is RISC, though it's dominant feature is its scalability (as you noted was a big plus with RISC). What we have is a segmented market. Once you have armies of Intel x86 and ARM programmers and tools in place, they won't get displaced easily.