I know there are some few threads around here, wich discuss the performance of the n770 and the n800 device, but there is no real performance comparison between Xscale an the omap2420 processor. Lets' start with this two:
Intel Xscale pxa270 @416Mhz:
Arch.: ARMv5TE
Wireless MMX(SIMD) Instruction Set
TI omap2420 @320Mhz:
Arch.: ARMv6
Core: ARM1136J(F)-S
FPU
DSP
IVA
How's the performance difference between these two CPU's ? (IVA not used on the N800)
I don't have any real stats to hand, but I'd say the OMAP2420 is equivalent to and possibly slightly faster than the higher clocked PXA270 due to the improved architecture of the ARM6 core. Also the OMAP2420 should wipe the floor with the PXA270 when running floating point intensive code.
The MMX instructions should give the PXA270 a boost in certain applications, however there are SIMD instructions available on OMAP2420/ARM6 as well but these don't seem to help significantly.
If the DSP, IVA and PowerVR MBX hardware could be better utilised on the N800 it would become an awesome pocket powerhouse. As it is, it's not anything special - a bit faster than the 770 is about the best you can say.
I know there are some few threads around here, wich discuss the performance of the n770 and the n800 device, but there is no real performance comparison between Xscale an the omap2420 processor. Lets' start with this two:
Intel Xscale pxa270 @416Mhz:
Arch.: ARMv5TE
Wireless MMX(SIMD) Instruction Set
TI omap2420 @320Mhz:
Arch.: ARMv6
Core: ARM1136J(F)-S
FPU
DSP
IVA
How's the performance difference between these two CPU's ? (IVA not used on the N800)