65 fps at both 1680x1050 and 1280x1024???

[five40]

So I just got a new 2005fpw and I decided to do the video stress test in CS:S. I ran the test and got 65 fps at 1280x1024 on my old 18" LCD. I swapped out to my 2005fpw, set the new resolution in the game and re-ran the test and I got the exact same score. How in the world does this happen? I've got a 9800 pro and a 2500xp@3200.

 

[SynthDude2001]

That sounds like being really CPU-limited, but I've got a similar CPU (XP1700@2.4GHz) and I used to get ~115fps at 1024x768. Are you sure vsync isn't on?

 

[Rage187]

youve hit the roof on one of your components.

 

[slash196]

Hey, count yourself lucky. My current rig chugs on BF1942...at 800x600 resolution. I don't need your pity. I just need money. Lots and lots of money.

 

[Peter]

So? None of the two LCD screens gets fed more than 60 frames per second. Rendering more than 60 frames per second makes no sense, since they're not getting displayed.

 

[nitromullet]

Isn't CS:S Stress Test capped at something like 80 FPS? If that's the case, an average of 65 FPS makes sense.

 

[five40]

Originally posted by: nitromullet
Isn't CS:S Stress Test capped at something like 80 FPS? If that's the case, an average of 65 FPS makes sense.

Maybe that's why. I'm completely happy with 65fps. I thought I was going to have to buy a new video card in order to play games at the native resolution of the 2005fpw. I'm very happy with the results, I was just curious as to why this might happen. Looks like I'll get to wait another video card/CPU generation before I have to upgrade. Woohoo for that.

 

[SynthDude2001]

Originally posted by: nitromullet
Isn't CS:S Stress Test capped at something like 80 FPS? If that's the case, an average of 65 FPS makes sense.

I'm pretty sure it's not capped; that would almost defeat the point of a video benchmark. Like I said above, I used to get around 115fps when I ran the test at 1024x768.

 

[five40]

Originally posted by: SynthDude2001

Originally posted by: nitromullet
Isn't CS:S Stress Test capped at something like 80 FPS? If that's the case, an average of 65 FPS makes sense.

I'm pretty sure it's not capped; that would almost defeat the point of a video benchmark. Like I said above, I used to get around 115fps when I ran the test at 1024x768.

What video card did you run to get those scores?

 

[SynthDude2001]

Originally posted by: five40

Originally posted by: SynthDude2001

Originally posted by: nitromullet
Isn't CS:S Stress Test capped at something like 80 FPS? If that's the case, an average of 65 FPS makes sense.

I'm pretty sure it's not capped; that would almost defeat the point of a video benchmark. Like I said above, I used to get around 115fps when I ran the test at 1024x768.

What video card did you run to get those scores?

6800GT at Ultra speeds.

I realize that would give me higher scores than you - but especialy with an "older" card, it doesn't make sense that the framerates would stay the same as you turn the resolution way up, unless the bottleneck is something else.

 

[five40]

Originally posted by: SynthDude2001

Originally posted by: five40

Originally posted by: SynthDude2001

Originally posted by: nitromullet
Isn't CS:S Stress Test capped at something like 80 FPS? If that's the case, an average of 65 FPS makes sense.

I'm pretty sure it's not capped; that would almost defeat the point of a video benchmark. Like I said above, I used to get around 115fps when I ran the test at 1024x768.

What video card did you run to get those scores?

6800GT at Ultra speeds.

I realize that would give me higher scores than you - but especialy with an "older" card, it doesn't make sense that the framerates would stay the same as you turn the resolution way up, unless the bottleneck is something else.

I'm thinking the CPU is bottlenecking everything. If it were the video card then the higher resolution would cause it to slow down.

 

[SynthDude2001]

Originally posted by: five40

Originally posted by: SynthDude2001

Originally posted by: five40

Originally posted by: SynthDude2001

Originally posted by: nitromullet
Isn't CS:S Stress Test capped at something like 80 FPS? If that's the case, an average of 65 FPS makes sense.

I'm pretty sure it's not capped; that would almost defeat the point of a video benchmark. Like I said above, I used to get around 115fps when I ran the test at 1024x768.

What video card did you run to get those scores?

6800GT at Ultra speeds.

I realize that would give me higher scores than you - but especialy with an "older" card, it doesn't make sense that the framerates would stay the same as you turn the resolution way up, unless the bottleneck is something else.

I'm thinking the CPU is bottlenecking everything. If it were the video card then the higher resolution would cause it to slow down.

Yeah, that is the most likely cause. Although like I said, my CPU isn't really that much faster than yours, so who knows... It could have been a Vsync issue, but then again the default refresh rate of the 2005FPW is 60Hz anyway (compared to 75 for some 1280x1024 panels), so in theory you shouldn't even be able to get 65fps if vsync is on.

 

[Peter]

I'll say it again: You don't "get" 65 fps. The graphics card may render them, but you don't get to see more than 60. Same for the folks with CRTs - with the refresh rate at 75 or 85, you're not going to see more than exactly this many frames, no matter whether your graphics card renders exactly those 75 or ten times as many.

 

[SynthDude2001]

Originally posted by: Peter
I'll say it again: You don't "get" 65 fps. The graphics card may render them, but you don't get to see more than 60. Same for the folks with CRTs - with the refresh rate at 75 or 85, you're not going to see more than exactly this many frames, no matter whether your graphics card renders exactly those 75 or ten times as many.

I think we all understand this; that's why Vsync exists. The point is that you can't accurately assess and/or compare (edit: mostly compare) video card performance when there's an artificial ceiling imposed on the frame rate. 🙂