Hello,I have the following piece of code:
So, after bx*16 bytes, we need to jump to another location in memory and read/store bx*16 bytes again, and so on.One possible ASM code for NEON to support the aforementioned function is the following (assuming that bx=by=4):
Throughput in this case is referring to the number of the same instruction thatcan begin execution on each cycle. The exact code layout is not particularlyimportant for large out-of-order cores like Neoverse N2 or Neoverse V2, so Iwould expect both arrangements to perform more or less the same. The bottleneckin such cores is instead usually any dependency chain between instructions, forexample in the case of the load instructions here the loads cannot beginexecution until the addresses x0 and x1 have been calculated.
Glad to hear the new code worked as expected!