Moving on to those examples, it's something of a mixed bag - John Carmack's engines have always had some awareness of multiple processor systems, so the Doom 3 engine-based Quake 4 was probably a natural target for dual-core optimisations. And, it has to be said, the results are impressive - Of course, you can't expect throwing any amount of processing power to help out in a GPU-limited situation, but where the CPU is even vaguely a limiting factor there is performance to be gained from enabling SMP in this game using the 1.05 patch.

Call of Duty 2 shows far less impressive results, and indeed a drop in performance at all but the highest resolutions, where it ekes out a tiny advantage - A rather bizarre state of affairs in all honesty. It would be interesting to see what happens with this title on Intel's dual-core and HyperThreading parts - The 1.01 patch is titled as an 'Intel update' after all!

From our testing here, the biggest cause of concern is the seeming incompatibility between the dual-core enhancements to NVIDIA's video card drivers and the improvements made to these game titles themselves - In Quake 4, with both NVIDIA's and the games dual-core enhancements turned on, there is a very large drop in performance, leaving you with an either/or situation as to which you have enabled. Obviously, this isn't a particularly satisfactory state of affairs, and is something that can hopefully be resolved as time goes on. Nobody is going to want to spend their time hacking through the registry depending on whether or not the game utilises multi-threading, so hopefully NVIDIA or the game developers will take some time to work out what's going wrong here and prevent it from happening in the future. At present I haven't repeated these tests using an ATI board and drivers, but will look into doing so to see if the recent CATALYST drivers with dual-core optimisations present the same kind of issues as we've seen here.