Average fps seems useless

[Ion]

After reading [H]'s 9600XT review(halo section) i began to think how useless average fps is when you in the thick of a battle.

Getting 60 average doesnt save you from getting 20 fps when you mostly needed. :tongue:

<A HREF="http://www.hardocp.com/article.html?art=NTMzLDM=" target="_new">http://www.hardocp.com/article.html?art=NTMzLDM=</A>

 

[dhlucke]

That's why I think everyone should benchmark for minimum fps.

_________________________________________
<font color=red>12 bit... The way games are meant to be played!</font color=red>

 

[rubikian]

Don't you think sometimes minimum fps would be bounded by cpu instead? If that's the case, it's pretty unfair for high end cards...

 

[phial]

why would it be? most of my games are bottlenecked by my TI4200 , and i only have a 2000+ athlon

-------


................

 

[WingDing]

I think whatever reflects real life performance is always going to be useful. And knowing how CPU and graphic cards scale together is critically important.

Barton 2500+, 512MB Corsair Platinum XMS 3200 CL2, Radeon 9700, WD Raptor 10,000 rpm S-ATA HDD, Asus A7V600, Enermax 460W SilentPlus PSU.

 

[Ion]

The newer bunch of games are heavily depended on graphic card performance especially true in FPS games. I think right now mostly CPU bounded games are strategy type due to AI processing.

 

[TKS]

I agree. So if I was a betting man, I'd bet that whoever designs a benchmark that that compares and contrasts CPU and VPU performance and uses a recordable performance indicator to output (say in ratio/performance indication) would probably become a millionaire.

----------
<b>I'm not normally a religious man, but if you're up there, save me, Superman! </b> <i>Homer Simpson</i>

 

[Ion]

Only problem is that the benchmark may quickly fallen to the dust once Nvidia figure out how to "tweak" it for FX cards... :tongue:

 

[TheGreatGrapeApe]

Yep, alot of s have been pushing for min. FPS in games.

Getting spikes is terrible. Som even post Hystograms that show the overall performance during a whole demo so you see the peaks and the valleys. They are harder to read but much more informative showing you when the different cards can't handle the stress and whether you might se moments of single digit fps.


- You need a licence to buy a gun, but they'll sell anyone a stamp <i>(or internet account)</i> ! <A HREF="http://www.redgreen.com" target="_new"><font color=green>RED</font color=green> <font color=red>GREEN</font color=red></A> GA to SK

 

[allanon1965]

i read somewhere that the brain can only handle up to 30 fps, anything over that is usless, if this is true, then what does the higher fps do for the game? curious.....

Fat, Drunk, and Stupid is no way to go through life Son.

 

[bandikoot]

IIRC, most things have motion blur, such as if you were to pause a single frame in an action sequence on a dvd, which kinda ties all the frames together. Computers don't do that unless programed so each frame is distinct and some people can tell that apart. Also, like has been mentioned before, if you have an average of 30fps put out by your benchmark that means you could have been bouncing between 10 and 50 throughout the course of the trial, and 10 fps looks jerky to just about anyone. Claiming 60fps generally ensures that the game did not spend much time at or below 30 fps.

 

[TheGreatGrapeApe]

Stop reading stuff! Makes your brain hurt, and my ears bleed! :tongue:

Since we've already covered this before, that is all.



- You need a licence to buy a gun, but they'll sell anyone a stamp <i>(or internet account)</i> ! <A HREF="http://www.redgreen.com" target="_new"><font color=green>RED</font color=green> <font color=red>GREEN</font color=red></A> GA to SK