AMDs Effect of a Cacheless Northbridge – Page 2

I decided to pick a hand full of the more popular benchmark programs. There are 100’s of programs and tools out there but I don’t have forever to run them all. 🙂

Below is the list in no particular order…

    • Super PI mod1.5 XS
    • 7-Zip 9.11 Beta
    • Cinebench r10 32 Bit
    • TMNations Racing
    • Aquamark 3
    • Fritz Chess Bench 4.2
    • Passmark PerformanceTest 6 (Summary Only)
  I tried to get accurate results in all benches. Unfortuntuately there are some benches that just didn’t want to be consistent. Instead of picking the peak number I picked a result between the min and max after about 3-10 runs (Depending on the bench). This method isn’t perfect but it should still provide a decent amount of accuracy comparing each Northbridge speed setting.

 

Everest Home 2.2
  This is a simple bench that I used to see what theretical difference that takes place between each NB setting. I checked Memory Latency, Memory Read, and Memory write performance.

 

 

  Although not by a huge amount, there are measurable differences in latency from raising the speed of the nb. From 1410 Mhz to 2820 Mhz there is a solid 14% decrease in latency.

 

 

Again, as a result of nb speed there are measurable differences in memory throughput for the read and write performance. Another solid 14% increase from 1410 to 2820.