Usually, it’s easy to benchmark CPU core and clock speed scaling, but with the 60 FPS limitation, it’s a little more complicated here. If we simply increase and decrease the number of CPU cores, we might see no difference at all if our clock speed is sufficient to keep the frame rate above 60 FPS. Thus, we’ll do the clock speed tests first to make sure we’re running lower than the vsync ceiling. We will disable CPU cores appropriately to make sure that if there are any differences, then they can be seen below the vsync limit.
We’ll run at a low resolution paired with the powerful Radeon HD 4850 video card to ensure that the graphics subsystem won’t interfere with the numbers.
Now that we have a method, let’s see how CPU clock speed affects Burnout Paradise:
We adjusted our clock speeds by lowering the CPU multiplier to keep everything else (such as memory speed) on an even keel. Note that the lowest multiplier available in the BIOS was 6x, which corresponded to the 1.8 GHz clock speed shown in the graph. Even at the dramatically-lowered clock speed, there seems to be little effect on game play. But it’s significant that the average speed is at least below the vsync limit of 60 FPS, which means we’re seeing a meaningful rate of maximum performance for that clock speed, and we are thus able to run our CPU core test.
The CPU core test shows us exactly what our recent How Many CPU Cores Do You Need? article suggests: this game utilizes up to three CPU cores efficiently and the biggest performance jump occurs between single and dual cores.