It depends on the architecture. I know that a register load on the Motorola 68k CPU will set the flags (Z for zero, N for negative) while a register load on the Intel x86 line doesn't change the flags and thus, a comparison needs to be made.
Both architectures, however, include special instructions to handle numeric loops (for instance for (i = S ; i != E ; i += step) that count downwards, but they have a slightly different ending condition (Intel ends on 0, Motorola ends on -1) and is really specialized on the Intel (it requires the use of the CX (ECX, RCX) register; on the Motorola you can use any of the data registers) so it makes it kind of hard to use in portable C code (your best bet---don't use the for index in the body of the loop and hope the compiler can transform it to use the specialized instructions).
If speed is really critical, you need to check the output from the compiler, and make sure you always measure twice (and keep the original code around as a reference when the hardware changes).
There is no need to compare - ORing the register with itself will suffice to set the Z flag. This is more concise on x86 - i think it's 2 bytes to OR a 32-bit register with itself, vs 3 for a compare with a (sign-extended) 8-bit immediate zero.
Both architectures, however, include special instructions to handle numeric loops (for instance for (i = S ; i != E ; i += step) that count downwards, but they have a slightly different ending condition (Intel ends on 0, Motorola ends on -1) and is really specialized on the Intel (it requires the use of the CX (ECX, RCX) register; on the Motorola you can use any of the data registers) so it makes it kind of hard to use in portable C code (your best bet---don't use the for index in the body of the loop and hope the compiler can transform it to use the specialized instructions).
If speed is really critical, you need to check the output from the compiler, and make sure you always measure twice (and keep the original code around as a reference when the hardware changes).