Originally Posted by RiverOfIce
Sigh. You can not buy a new n450. The chip has been discontinued. So you are looking at per watt/can not buy. Secondly. Any software has to be optimized for the cpu. Using the same software just proves that gentoo runs better on the arm system. Bad software on a good cpu will not make the cpu work well.
As for the current atom processors. The slowest atom processor is faster then the fastest and newest arm. http://liliputing.com/2012/02/fastest-arm-chips-are-comparable-to-intels-slowest-atom-chips.html
But it really comes down to coding, better coding equals faster scores.
AS for the "troll" comment. I am pretty sure you just broke the rules of the rules.
Just did a google search there are still some oem's out there selling the n450 on the low end scale. So ignore the can not buy comment.
Your ... (omitted), when you pass compiler flags through GCC as -march=atom anything that CPU and do will be used. The ENTIRE opsys is built CPU specific in that sense, unless they just used generic 32bit binaries/builds. But since they specifically showed you the flags, it's atom specific. Plus Atom hasn't changed much in it's lifetime, you still get about the same watt/performance you did back then. Not to mention the link you posted specifies the ARM at 1GHz and the Atom at 1.6GHz, that's a bit off isn't it? Though you don't have to review them at equal clocks, but if you want to say price per watt used and do an equal comparison you probably should.
You didn't make your statement sound any better, seriously you need to know how code is optimized at the compiler level. -march=atom is all you need if your building the software from source. So that exact same software isn't going to have much of Arm/Atom specific "optimizations". Here, let me make it simple for you:
The first and most important option is -march. This tells the compiler what code it should produce for your processor architecture (or arch); it
says that it should produce code for a certain kind of CPU. Different CPUs have different capabilities, support different instruction sets, and
have different ways of executing code. The -march flag will instruct the compiler to produce code specifically for your CPU, with all its
capabilities, features, instruction sets, quirks, and so on.
I tread through crap like this on a regular basis compiling my own software for those programs that could use a little fat cutting. Even though I run a binary based distro (currently Fedora 17) I still dabble in building my own crap. The biggest area is the kernel, so my kernel is arch specific and I can add my own patches.
 And yes, there aren't any real atom specific "optimizations" but things like (from my cat /proc/cpuinfo): "fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm 3dnowext 3dnow constant_tsc rep_good nopl nonstop_tsc extd_apicid pni monitor cx16 popcnt lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt hw_pstate npt lbrv svm_lock nrip_save" will be taken into account. If the code has the ability to optimize for SSE/SSE2 then -march=atom will find it or whatever flags that specific architecture can handle.Edited by mushroomboy - 9/5/12 at 9:41pm