Ok, enough with synthetics, here's 
  an older game, MDK2. Let's see what it thinks as it is good at taxing a computers 
  subsystem.
 
  
  
    | MDK2 640x480-16bit | 
  
    | Frequency | FPS | 
  
    | 170 MHz | 204.0 | 
  
    | 150 MHz | 192.6 | 
  
    |  | 
  
    | MDK2 
      1024x768-32bit      | 
	 
    | Frequency | FPS | 
  
    | 170 MHz | 193.5 | 
  
    | 150 MHz | 185.7 | 
          
                  
  
The difference wasn't noticeable while benchmarking; 
however it's still almost 6% on the first set of benchmarks at 16bit. As the 
resolution is turned up, there isn't any noticeable difference, but the numbers still favor the higher bus 
  speeds
Quake III Arena Benchmarks 
             
        
Quake III has always craved high bus speeds. Judging by the other previous 
  benchmarks, it's obvious that the system with the faster bus will win, but by 
  how much?
  
  
    | Quake III Arena: Fastest Setting Demo001 | 
  
    | Frequency | FPS | 
  
    | 170 MHz | 238.1 | 
  
    | 150 MHz | 220.8 | 
  
    |  | 
  
    | Quake III Arena: Fastest Setting 
      nV15Demo | 
	 
    | Frequency | FPS | 
  
    | 170 MHz | 68.6 | 
  
    | 150 MHz | 64.0 | 
At these settings, again, the faster bus wins out. Not by much, but still, it's "free" 
  performance. Let's take a look at what happens when the resolution is turned 
  up to 1024x768.
 
  
  
    | Quake III Arena HQ: Fastest Setting 
      Demo001 | 
  
    | Frequency | FPS | 
  
    | 170 MHz | 204.6 | 
  
    | 150 MHz | 195.5 | 
  
    |  | 
  
    | Quake III Arena HQ: Fastest Setting 
      nV15Demo | 
  
    | Frequency | FPS | 
  
    | 170 MHz | 53.0 | 
  
    | 150 MHz | 49.8 | 
  
  
    | Quake III Arena MAX: Fastest Setting 
      Demo001 | 
  
    | Frequency | FPS | 
  
    | 170 MHz | 161.7 | 
  
    | 150 MHz | 157.1 | 
  
    |  | 
  
    | Quake III Arena MAX: Fastest Setting 
      nV15Demo | 
  
    | Frequency | FPS | 
  
    | 170 MHz | 50.0 | 
  
    | 150 MHz | 47.0 | 
Again, I wasn't able to see the performance difference, but the numbers don't 
  lie. Hey, it's free performance, why not take it?