
Supreme Commander is one of the first games to feature out-of-the-box multi-core processor support. Will a quad-core CPU really show better gameplay? Real world gameplay results plus Vista versus XP performance findings with SupCom.
For evaluation we are using an EVGA nForce 680i SLI motherboard. We are using an Intel Core 2 Quad QX6700 2.66 GHz processor and 2GB of Corsair XMS2 Dominator CM2X1024-8888C4D at 4-4-4-12 1T. We are using the latest chipset drivers available and the latest BIOS at time of evaluation.

Today in Supreme Commander we are comparing single-core processors, to dual-core processor, to quad-core processor performance. Therefore our video card we are using will stay the same. We have chosen two BFGTech GeForce 8800 GTX video cards in SLI to remove the GPU bottleneck from the game experience.
We are running Windows Vista Ultimate 32-bit as well as Windows XP Pro SP2 32-bit. Windows Vista should support multi-core CPUs better than Windows XP. We will see if this makes any differences in the following comparisons.
For Windows Vista we are using the latest beta Windows Vista driver at NZone, ForceWare 101.41 which does support SLI in Windows Vista. Note that we also tried the ForceWare 100.65 drivers on NVIDIA.com and we saw the exact same performance between them in Supreme Commander.
For Windows XP we are using the latest official driver at NVIDIA.com, 97.92 WHQL.

With the EVGA 680i motherboard we can disable each core as seen above. Therefore the comparison is done with the same CPU and system configuration; we have just disabled cores to represent “Dual-Core” and “Single-Core” processors.
First we are going to run our “highest-playable” tests on a standard aspect ratio 19” CRT and then on the Dell 3007WFP 30” Widescreen LCD. On each page you will first have Windows Vista performance, and if you scroll down you will see Windows XP Pro SP2 performance. We will then follow with a suite of “apples-to-apples” tests and CPU utilization.
Please be aware we test a bit differently from what is the norm in the industry. We concentrate on examining the real-world gameplay that is provided. The following testing does not represent “canned benchmarks” and “canned demos” you will find elsewhere. The following article represents countless hours of real gameplay scenarios played out over and over to validate results, just like you would get if you were playing the game at home.