6 replies to this topic

[Al Bert]

Hi,

since this is still (open) beta, I think we are still here to test things out...now that i have some more data (see here, also for more system data http://mwomercs.com/...ost__p__1507770, i can show a comparison for FPS on the various maps:

EDIT: updated version (4th dec)



Overall, statistic is not very good (e.g. I'd like to have more games in each mode but there is this...uhm...real life going on somewhere ).

What i find "interesting" in a technical sense is the fact that River City shows the largest spread with > 15 FPS for lowest and highest possible setting (on my system). Also, the FPS on this map is ~5 FPS lower compared to all other maps in my settings "high" and "medium".

What i also observed with low settings is that in-match detail is so low, I sometimes don't see mechs in large distance while having them locked with R, so i can simply fire in the middle of the big square. Also terrain many times look awfully flat (which i think is "ok" in the sense of normal for low settings).

I also tested on a very low end system with a GT9600M card and a dual core - with lowest setting, i could not make it above 10-12 FPS in the best case (Windows performance index is 5.0 on that system).

I am anxious to see the numbers in a week from now.

Regards,
Al Bert

Edited by Al Bert, 04 December 2012 - 07:43 AM.

[EvilCow]

Thanks for data, it matches my experience on my low end dual core system. Overall good performance everywhere but Forest Colony.

Edited by EvilCow, 02 December 2012 - 01:41 PM.

[p4r4g0n]

Thanks for going to the trouble of compiling and sharing your data. Find it interesting the lowest fps are on both frozen colony snow and frozen city night.

I had the impression that I was getting best fps from frozen city night but I wouldn't bet my life on it though.

Did you do a client restart for every data point to avoid the variability caused by the memory leak bug?

Edited by p4r4g0n, 02 December 2012 - 01:42 PM.

[Al Bert]

p4r4g0n, on 02 December 2012 - 01:41 PM, said:

Did you do a client restart for every data point to avoid the variability caused by the memory leak bug?

Actually i did in the beginning but did not find any significance to it. Also Motion blur ON/OFF did not change anything. If at all, the daytime (probably related to the server load) makes ~3 FPS difference (less in my evening, GMT).

[Phelan Anastasius Kerensky]

Al Bert, on 02 December 2012 - 01:35 PM, said:

I also tested on a very low end system with a GT9600M card and a dual core - with lowest setting, i could not make it above 10-12 FPS in the best case (Windows performance index is 5.0 on that system).

Same here. As soon as i go into close combat it drops to 4-6 FPS. AMD X2 Dual Core with a GTX 260 and 4Gig RAM.

[Xenoid]

The worst map for me in regards to fps is river city. All other maps are equal just about.

[EnigmaNL]

Nice graph, it's the same kind of fps I am getting at those settings. It annoys me to see this low performance in this game. Crysis 2 (also on Cryengine 3) run a whole lot better on my system.