• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Interpretting benchmark help!

Soldato
Joined
7 Aug 2012
Posts
2,652
Reading July issue of PC Gamer and there is a section on multi-gpu's. Anyway, the benchmarks aren't particularly in depth, but do state they are taken at highest setting, with 4xMSAA and on a 2560x1600 panel.

I am putting together a 7970 Xfire rig and just looking to try correlate what I can roughly expect, although currently game at 1920x1080 (perhaps looking to go 120Hz, or larger res in near future).

Anyway, looking at it on a basic level is it correct to assume that I can work out what the results would be at 1920x1080 simply by deducing that a 2560x1600 panel has 1.9753 the amount of pixels and multiplying the fps in the benchmarks by that figure? Obviously I understand it wont be exact, but would that be a rough way of looking at things, or do gpu results not scale at all like that?
 
Take this with a pinch of salt, but these were compiled ages ago when Gregster and Spoffle were arguing over scalability between resolutions, pixels scale linearly with load etc.

In short, going from 1080 to 1600 even though the pixel count is 97% more the load on the GPU was only an average of 44% more, whereas going from 1600 to 5760 the pixel count is 51% more but the average load was also 47% more.


The source data was from Techpowerup, and showed some anomalies of testing for the 7970s hence taking the figures from Titan data.


 
basically, no - I certainly didn't have half the FPS when I went from a 1920 to a 2560 monitor... more like 1.4-1.5 in say BF3
results will vary game by game and depending on settings

why not just google for 7970 crossfire results at 1920x1080?
 
So looking at that very briefly, would it be correct to assume that the FPS would drop by 47% going from 1080 to 1600, which is pretty close to what one would reasonably expect?

Cheers for the info though, will look at it in more detail later.
 
basically, no - I certainly didn't have half the FPS when I went from a 1920 to a 2560 monitor... more like 1.4-1.5 in say BF3
results will vary game by game and depending on settings

why not just google for 7970 crossfire results at 1920x1080?

Aye, will have a wee look for the results at some stage lol. Just wondered if results could be scaled using that method based on pixel count. It would appear not as straightforward however.

Thanks.
 
Back
Top Bottom