Tom's Hardware has helped me choose many great system components over the years and I want to express my gratitude for that. I am writing this because I believe there is a better way to benchmark CPU's using Supreme Commander.
When someone is playing an RTS like Supreme Commander there are two factors to be considered, FPS and Max Game Speed. For most people playing an RTS, the max game speed is as important or more important then FPS. If your game is running at -4 it requires four hours to play one hour of game time.
The issue I have with only testing FPS is that it is possible for the FPS to improve and the max game speed to decrees. I experienced this first hand when I went from an AMD X2 5000+ Black edition to a phenom 9500 my frame rate went up by about 15-20% while my max game speed decreased 10-15%. In addition the max game speed only seem to be affected by CPU speed. Changing graphics cards has little to no effect on max game speed.
I am suggesting that you also bench the max game speed in the following manner. On a large scale map create the max number of AI's and record a demo (say 30-40 minutes). then start the demo at max speed (10x) and record the time it takes to complete on each system. I would also like to see the lowest game speed each system reached during play back. max game speed can be displayed by typing ren_shownetworkstats in the console.
Concluding notes:
I searched the net for anyone testing game speed and did find that hardwarecanucks.com was doing something along these lines.
http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/11390-intel-core-i7-nehalem-920-940-965-xe-processor-review-17.html
I would be interested to know other forum readers thoughts and what Tom's Hardware think of benchmarking this way.
When someone is playing an RTS like Supreme Commander there are two factors to be considered, FPS and Max Game Speed. For most people playing an RTS, the max game speed is as important or more important then FPS. If your game is running at -4 it requires four hours to play one hour of game time.
The issue I have with only testing FPS is that it is possible for the FPS to improve and the max game speed to decrees. I experienced this first hand when I went from an AMD X2 5000+ Black edition to a phenom 9500 my frame rate went up by about 15-20% while my max game speed decreased 10-15%. In addition the max game speed only seem to be affected by CPU speed. Changing graphics cards has little to no effect on max game speed.
I am suggesting that you also bench the max game speed in the following manner. On a large scale map create the max number of AI's and record a demo (say 30-40 minutes). then start the demo at max speed (10x) and record the time it takes to complete on each system. I would also like to see the lowest game speed each system reached during play back. max game speed can be displayed by typing ren_shownetworkstats in the console.
Concluding notes:
I searched the net for anyone testing game speed and did find that hardwarecanucks.com was doing something along these lines.
http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/11390-intel-core-i7-nehalem-920-940-965-xe-processor-review-17.html
I would be interested to know other forum readers thoughts and what Tom's Hardware think of benchmarking this way.
Facebook
Twitter
Instagram