Author: Ratko V Tomic
Date: 14:47:54 07/27/03
Go up one level in this thread
In the case of constant m the XOR-method needs 1 immediate 4 byte value while the OR-method needs 2 immedate values, adding 4 bytes to the code. Depending on the context and the CPU pipeline and cache sizes & states, this would have some degrading effect. The parallelization in the OR-method may perhaps offset it, depending on whether other orthogonal instructions can be interleaved. One would probably need to test it in the specific context (and CPU) to see which factor has greater effect. In the case of variable m, the XOR-method should always be faster.
This page took 0.02 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.