Why we need a benchmarking test in NS2
nsguy
Join Date: 2010-01-03 Member: 69869Members
Join Date: 2010-01-03 Member: 69869Members
For those who don't know what benchmarking is, take Counter-Strike: Source, on the main menu there is a stress test option that once clicked, runs through a scripted scene. It monitors statistics about how your computer is performing in rendering the scenes. (3DMark is another example.)
Reasons for it:
1. It shouldn't be difficult to implement considering UWE created "film making" software so that they could script scenes.
2. It will allow us to compare performance in a more scientific fashion; a "laboratory" setting.
3. It will allow us to change the video settings in the game and determine which combinations perform best.
4. In a time when lots of people are complaining about performance and UWE is spending a lot of time trying to optimise the game, this would be most useful.
Reasons for it:
1. It shouldn't be difficult to implement considering UWE created "film making" software so that they could script scenes.
2. It will allow us to compare performance in a more scientific fashion; a "laboratory" setting.
3. It will allow us to change the video settings in the game and determine which combinations perform best.
4. In a time when lots of people are complaining about performance and UWE is spending a lot of time trying to optimise the game, this would be most useful.
Comments
The info could be submitted (user's option) to UWE to see how different hardware is running the game. Could definitely be helpful for further optimizations
Either way I agree. I know UWE has their own personal system for getting stats but that does not tell the users anything. A benchmark test would really show if a patch increases or decreases performance as well as compares settings so people could share results to show which hardware works bests. Instead of people just feeling how the game runs which does not really tell us much sadly. In addition to letting people choose the best settings.
I see what you did there :)
I agree, a standardised benchmark would not only help UWE to optimise performance with the extra data gathered but would also help individual users tweak their own systems to what works best.
you can already record demos
someone just needs to record a run specifically with benchmarking in mind and call it the benchmark demo.
There are also no ways to skip segments of the demo, "keyframes" or any of that.
If the demo recording was improved, it could easily be used to benchmark...