Resident Evil 5 - 3D Benchmark

Dutch Guy · 24 Jul 2009 at 18:29

[email protected]
GTX280@700/1400/2400
1900x1200 4xAA 16xAF
WinXP 32bit

Phatjayunoi · 24 Jul 2009 at 18:34

Will_3rd said:
Running not too shabby, Everything on maximum settings, 3.0Ghz overclock.

Not too shabby indeed.

It would run just like the console versions, framerate wise, but only with more detail and at least in a higher resolution

.

Dano · 24 Jul 2009 at 19:31

Not read the rest of the thread yet but my initial tests indicate XPx86 to be roughly 10% better frame rates than W7x64 in DX9, using DX10 on W7x64 actually increases fps. I need to verify my driver versions and recheck though.

arc@css · 24 Jul 2009 at 20:12

Guess my previous results were due to having not only all settings in game set to the highest but in the ATi control panel too. This time I got 88.3fps average compared to 15 - 20.

Reckon I can tweak it some more to push it a bit higher still.

edit : 89.4 when also enabling vsync in control panel, forgot to increase fan speed though. Whether that will help much I don't know.

Dark_Angel · 24 Jul 2009 at 22:36

What does the console versions run at? I haven't read anything to tell me the frame rate, but my friend and I was running the demo to show him the PC version (since we both earn own the PS3 version) and we think the console versions run at 30FPS? we didn't have them to compare side by side, so we was basically winging it from memory that the PC versions seemed smoother.

Phatjayunoi · 24 Jul 2009 at 22:42

Dark_Angel said:
What does the console versions run at? I haven't read anything to tell me the frame rate, but my friend and I was running the demo to show him the PC version (since we both earn own the PS3 version) and we think the console versions run at 30FPS? we didn't have them to compare side by side, so we was basically winging it from memory that the PC versions seemed smoother.

Yeah 360/PS3 versions are set at 30FPS.

The PC benchmark definately shows it can run a lot smoother, or could also be set to be the same as the consoles, only difference is that you could run higher details and higher resolutions for the same framerate depending on your system.

Shame that seeing this benchmark running at 60FPS+ gives me really bad motion sickness.

Dark_Angel · 24 Jul 2009 at 23:28

Ah ok then, we were right

snowdog · 24 Jul 2009 at 23:38

ViRuS2k said:
so for him to be getting 90fps avg on 1 260 even 2x 260s wouldnt get that not at 1920x1200.

Yes they would, one gtx260 gives me 74 fps at 1920x1080 with 4x AA and 16X af in dx10.

If he'd run at 1280x1024, he'd easily get 90 fps with 1 gtx260, or in dx9, or running no AA...

thepigf@rmer · 25 Jul 2009 at 20:42

Tested in DX9, dodgy results:

Had a quick Google:

Renaming dx9 exe to fear2.exe fixed my fps.

Ta-daa!!!!!11!

.......maybe I should have stayed with Nvidia

Rapido! · 26 Jul 2009 at 05:19

ViRuS2k said:
cid is probably at 1280x1024 res LOL

i get 94fps avg @ 1920x1200 4x aa all max settings but im useing 2x 4890s uber clocked.

so for him to be getting 90fps avg on 1 260 even 2x 260s wouldnt get that not at 1920x1200.

Wasn't going to d/l this, but this post here set the curiosity cat amongst my metaphorical pigeons, oh yes it did.

So here we go; ran the maxed DX10 bench, motion blur on, 4XAA at 1680x1050 resolution on a single (albeit very highly overclocked) 260-216, fed by a Core i7 at 3.33 GHZ.

92.5 frames. Allowing for, say, a 10% performance hit, my system as it is would still be hitting an average in the mid 80s up at 1920 res.

Conclusion? This is a very Nvidia-friendly game.

ViRuS2k · 26 Jul 2009 at 05:35

Well that is extremly weird.
i mean i get worse results on area3... 1920x1200 4xaa

i mean 1 4890 @ 1000/1150 should theoreticly be far faster than a 260 but in this demo its not well weird and frankly nvidia based game or ati based game what ever it runs better on it clearly has massive issues with ATI cards.

dirtydog · 26 Jul 2009 at 08:58

thepigf@rmer said:
.......maybe I should have (sic) stayed with Nvidia

Yes because I'm sure averaging 169 fps is going to be a real bummer for you?

HouR · 26 Jul 2009 at 09:20

Will_3rd said:
Running not too shabby, Everything on maximum settings, 3.0Ghz overclock.

sorry to quote a pic but I like to see the comparison between the 8800 and 3870.

how much AA were you running with the above benchmark?

everything maxed with 2xAA I get 33 average and the pic below is everything maxed but no AA

obviously there is a slight difference in res and my overclock is in sig

dirtydog · 26 Jul 2009 at 09:58

That seems like a great score for the 3870, which AFAIK is similar performance to my card. You got a massive boost by turning AA off, but I don't.

Me with 2xAA:

dirtydog said:

And no AA:

Chong Warrior · 26 Jul 2009 at 14:00

Just been running a few quick tests as I'm about to try updating to the latest 190 drivers from the old 181.22's. Got some weird results!

E8400 @ 3.8ghz
260GTX 650/1150/1350
1920x1080
Fixed Benchmark

DirectX 10 results
No AA, Forced AFx16 = 57.7 fps
x4 AA, Forced AFx16 = 57.7 fps
x8 AA, Forced AFx16 = 54.2 fps
x16Q AA, Forced AFx16 = 53.1 fps

DirectX 9c results
No AA, Forced AFx16 = 52.8 fps
x8 AA, Forced AFx16 = 52.2 fps
x16Q AA, Forced AFx16 = 51.4 fps

Going to run the same tests again with the new drivers.

Variable Benchmark

DirectX 10 results
No AA, Forced AFx16 =
Part 1 = 131.3 fps
Part 2 = 89.1 fps
Part 3 = 85.2 fps
Part 4 = 94.0 fps
97.7 fps Rating S

Screenshot below is with 8xAA and 16xAF. Big performance hit unlike in the fixed benchmark which suggests something isn't working properly in that test...

dirtydog · 26 Jul 2009 at 14:41

Going from the 186 to the 190 drivers made no difference on my machine.

I'm curious how some people with what seem to be comparable machines to me (eg. 8800 GTS or HD 3870 and on only dual cores) are in some cases getting significantly better scores.

Rapido! · 26 Jul 2009 at 18:48

ViRuS2k said:
Well that is extremly weird.
i mean i get worse results on area3... 1920x1200 4xaa

i mean 1 4890 @ 1000/1150 should theoreticly be far faster than a 260 but in this demo its not well weird and frankly nvidia based game or ati based game what ever it runs better on it clearly has massive issues with ATI cards.

Nvidia got the jump on ATi with the drivers on this one; I'd imagine it'll all be ironed out on the ATI side by release date.

Also, you're underestimating the 260 there. My 216 core at it's factory overclocked speed was already nip and tuck with a GTX 280, and I run it at significantly higher clocks than that again. I actually used to have two in SLI but the other one has since passed on. To the other side. Into death.

Chong Warrior · 27 Jul 2009 at 19:46

After updating to the 190.38's from the 181.22's (results above) I ran 4 of the rests in DirectX 10 mode.

Absolutely zero difference, not even 0.1fps different in the fixed benchmark.

In the variable the 190's were actually 1 fps slower.

Disappointing considering that's a pretty big jump in driver release dates.

arc@css · 28 Jul 2009 at 00:48

The new 9.7 ATi drivers doesn't seem to have improved my fps average, were good anyway, but was hoping / expecting more performance.

Muse_Cubed · 28 Jul 2009 at 12:11

I got 26.6fps...1920x1200, VSync, 4xAA, everything on high (DX10).

I'm on a Q9450 @3.4Ghz and Radeon 4870.

It's a bit of a departure from the other games...fear through swarming, but I'm not convinced about the copious amounts of ammo. Resi always did a good job of making you feel really cautious about every bullet you spent. It's not quite as creepy (I think the monks in Resi 4 take the biscuit there, I always bricked myself when I heard them chanting behind me). Maybe it'll get back to the more traditional style later in the game. Now, if only I could use my Wii Remote for it...