Jump to content
Sign in to follow this  
MartinCrank

Video only benchmark

Recommended Posts

AFAIK when you load some mission and have FPS shown in something like FRAPS, you can click pause at any time and game does not go anywhere (no menu loaded, its on screen), although your FPS go up by 10-20 when paused, thats because CPU is not used so if you change graphics settings and click "hide" you will see max FPS your video card can output (because CPU is not used on pause). You can see how much your system bttlenecked because of CPU this way.

Is that correct?

Share this post


Link to post
Share on other sites

I would say yes. Although your cpu is always in use, even if the game is paused it'll still use cpu time to render the scene. So fraps can only show you a part of the problem. But as long as the fps goes up when pausing I'd say it's pretty safe to assume your cpu is bottlenecking.

Share this post


Link to post
Share on other sites

In a way yes but consider that for mission A the bottleneck could be the CPU while in mission B it's the gfx card. Visual quality settings will make a difference in the balance too, obviously.

Also, the term bottleneck isn't likely the right one in this case, as any cpu can't run the cycle with zero latency and you would probably get better fps with a faster gfx card anyhow.

Share this post


Link to post
Share on other sites
Also, the term bottleneck isn't likely the right one in this case, as any cpu can't run the cycle with zero latency and you would probably get better fps with a faster gfx card anyhow.

Could you explain this in greater details?

Share this post


Link to post
Share on other sites

I'm probably talking out of my ass here but what I meant to say that there's these two tasks to be done and they can't be done at the same time thus, making either one run faster would increase the fps. Let's suppose it could be simplified like this:

1. do all the simulation, read controls, run animations, physics, AI, whatnot

2. do all the rendering. go back to 1

Now, let's say task1 takes 20ms and task2 takes 40ms. With your method and this scenario it would appear that the cpu makes your game run 30% slower, while the major reason for bad fps would still be the gfx card.

Those values are of course purely fictional and might not actually be how the engine runs at all.

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
Sign in to follow this  

×