Quote:
Originally Posted by StainlessS
Any general purpose 32bit -> 64bit Exe conversion is likely to be about 5% to 15% slower (due to double size of pointers and also perhaps array indexes), though video plugs may be faster as more ram can be used for cache.
|
I think the slowdown due to double pointer sizes is neglible. In x64 there are much more general purpose and XMM/YMM registers are available, compiled code can be more efficient because processor registers are used instead of temporary variables swapped from/to memory.