Overclock.net banner
1 - 9 of 9 Posts

· Banned
Joined
·
9,061 Posts
Discussion Starter · #1 ·
Figured a I'd do another review, this time on Mafia II with PhysX set on high.

I also forced some higher settings through the NV Control Panel in order to improve the overall image quality in the game.

Test bench consisted of a i5-2500k @ 4.8 GHz, with 4GB of DDR3 @ 1900 running 8-9-8-24-1T in single channel mode (qq). Stock settings for the GTX 470 are 607 Core, 1214 Shader, and 1701Mhz Memory (3402Mhz effective). Overclocked settings for the GTX 470s are 850 Core, 1700 Shader, and 2200Mhz Memory (4400 effective).

The 9800GT was constantly at 810 core, 1893 shader, 918Mhz Memory (1836 effective) - Of note here is the 9800GT cards don't go exactly where you tell them, so these clocks could be slightly higher or lower depending on where the closest actual clock rates are.

Game Benched was Mafia II, using the IG benchmarking tool without using it to record the fps. FRAPS was used to record min/max/avg fps and that data was input into Excel to create a graph. The bench started once the fat guy got out of his card, and ended just as the screen began to fail after the windows are blown out.

images?q=tbn:ANd9GcSKBOmk6Lq0_WP61ROz5Jc-HX-oeYblz0eQne4TVwNImgiihIccoA


In game settings were set to their highest, forced graphical settings are as follows:

c1298058.png


And without further delay, here are the results:

167837ae.png


Later on I will do a similar review as this with Metro 2033, I hope these results help you. It should be noted that when the 9800GT isn't the most powerful card to use, however the Mafia II benchmark is excessive with its PhysX usage, and doesn't directly reflect the actual IG performance when using the 9800GT.

ninja.gif
 

· Premium Member
Joined
·
6,028 Posts
I'm a bit surprised..with the same settings i get ave. 62FPS on single 570
mafia2e.th.jpg

Uploaded with ImageShack.us
 

· Banned
Joined
·
9,061 Posts
Discussion Starter · #3 ·
Mafia 2 will give different results - they closed the benchmark thread for that reason.

It seems we're also using different versions, as mine is the Demo version and shows more info than yours does.

If you have the Demo version I reset my control panel to default and just used the IG settings, everything on and at max, no PhsyX single card at 850 core.

9173e600.jpg
 

· Premium Member
Joined
·
1,824 Posts
@BallaTheFeared

Great work
thumb.gif


Are you sure that those forced graphical settings are working, should be nice to see a comparison with in game highest vs forced.
 

· Premium Member
Joined
·
6,028 Posts
more info? the only difference is SP1 pack that i don't have and yes I'm running benchmark that came with the game ,will run tomorrow with no PhsyX and see how they stack up
Thanks for the review
meeeh who want's to wait..here all max, no PhsyX single card at 860 core.

mafia22b.jpg
 

· Banned
Joined
·
9,061 Posts
Discussion Starter · #7 ·
Quote:
Originally Posted by coolhandluke41;13094001
more info? the only difference is SP1 pack that i don't have and yes I'm running benchmark that came with the game ,will run tomorrow with no PhsyX and see how they stack up
Thanks for the review
meeeh who want's to wait..here all max, no PhsyX single card at 860 core.

mafia22b.jpg
Yeah don't worry too much about my results, the game seems to randomly select what settings it wants to use on its own without them before forced in the NV Control Panel. Often it seems to default back to the default settings, which explains why your 570 is nearly 50% faster when it shouldn't be.

The thread wasn't meant as a benchmark comparison thread, it was more meant to show 470s scaling and the 9800GT performance both with a single card and in SLI; in a benchmark that goes beyond any other game as far as PhysX demand goes (even further than the actual game).
 
1 - 9 of 9 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top