R600vsG80 Benchmark Comparisons From Any Game Welcome!

[Xion X2]

Obie, what are you getting on that 2nd temp reading w/ your overclock? I'm thinking this stock heat-spreader isn't going to cut it and that I'll have to invest in some ramchips soon.

[Obie]

turn the blower on to 70% that way it will keep things around 50C (at least for me).
Also, 1.20V is hardly scratching the surface here. It's more like 1.238V to 1.250V at 900 GPU.
With Mem at 880 and higher I found increasing the MVDDC one notch above 2.20V helps increase the min frame rates. Not sure about MVDDQ yet but from what I do know increase it a notch helps bring some balance to the MVDDC. Once things are setup rung Unigine at 1680x1050 4x/16x with everything enabled.

[Bman212121]

Got the HDR feature working right on the Catalyst 8.38s. Not sure what was happening before.. maybe adjusting an in-game option was screwing it up. But it's working now. Performance is relatively the same.. consistent 50+FPS w/ the hood cam. Settings same as before: 1680x1050, 4xAA, 8xAF:



Not too shabby. Is the card running at the numbers in your sig? Also, does this card has the occasional stuttering/slowdown or is it uneffected?

I vsync to 60 at 1920 x 1200 4x AA, 16x AF no HDR and it never backs off there except when you are on the bottom left of the map in all of those houses, then it might drop into the 50's. (Minus the stuttering of course)

[Xion X2]

Not too shabby. Is the card running at the numbers in your sig? Also, does this card has the occasional stuttering/slowdown or is it uneffected?

I vsync to 60 at 1920 x 1200 4x AA, 16x AF no HDR and it never backs off there except when you are on the bottom left of the map in all of those houses, then it might drop into the 50's. (Minus the stuttering of course)

Clocks were at 875||975 but I'm not seeing much of a difference on Test Drive w/ my overclock. Seems to run about the same at stock speeds. It's showing less of an increase so far than any other game I've benched.

It'll hitch every now and then, but not terribly bad. Might do it for a split-second every 30-45 seconds or so when it appears to be loading in new areas. I don't get any "slowdown" like you talk about.. it runs just the way you see in my pics. Consistent 50+FPS. W/ the hood cam, it stays ~55 most of the time; with the in-car cam, ~50.

W/ HDR disabled it stays pinned at my refresh rate of 60. Haven't ran it w/ Vsync off, so I have no idea what it would reach to w/o HDR.

I found out that turning vsync on/off in-game was what was screwing the HDR up before. All you have to do is deselect HDR and then reselect it and it comes back fine.

[Bman212121]

Clocks were at 875||975 but I'm not seeing much of a difference on Test Drive w/ my overclock. Seems to run about the same at stock speeds. It's showing less of an increase so far than any other game I've benched.

It'll hitch every now and then, but not terribly bad. Might do it for a split-second every 30-45 seconds or so when it appears to be loading in new areas. I don't get any "slowdown" like you talk about.. it runs just the way you see in my pics. Consistent 50+FPS. W/ the hood cam, it stays ~55 most of the time; with the in-car cam, ~50.

W/ HDR disabled it stays pinned at my refresh rate of 60. Haven't ran it w/ Vsync off, so I have no idea what it would reach to w/o HDR.

I found out that turning vsync on/off in-game was what was screwing the HDR up before. All you have to do is deselect HDR and then reselect it and it comes back fine.

I just played around a little with it, and at your settings I'm getting right around 55-60 fps having vsync off. What's interesting is that HDR on to HDR off at 1680 x 1050 is that it only take about a 5 fps hit. When I turn it on at 1920 x 1200, it cuts the frames right in half. That is probably why I turned it off way back when since I'm only running mid 30's with HDR on.

FYI my card is OC'ed to 648 / 2000 while I'm doing these tests, and I've been using 3rd person. I'll see if there is any difference between views or not.

The slowdown is basically the hitching. I didn't mean like what FEAR does where it goes from 90 to 9 fps and stays there, but it will drop down for a second when it hitches. I really hope they can fix that in the patch.

EDIT: There is actually quite a bit of difference between cams. 3rd person is at about the same as the ****pit view, but the hood cam is about 10fps more, and the road cam is around 12 fps more. The close up 3rd person view gives over 20 fps more than regular 3rd person, too bad you can't see what is in front of you. ;)

[Xion X2]

I just played around a little with it, and at your settings I'm getting right around 55-60 fps having vsync off.

At 1680x1050?

FYI my card is OC'ed to 648 / 2000 while I'm doing these tests, and I've been using 3rd person. I'll see if there is any difference between views or not.

Try using the hood cam if you can. Mine stays right around 55fps. I also still have my Conroe clocked at stock speeds (2.4GHz) for right now until I can customize a new BIOS for this board, so it should be fairly comparable to your FX @ 2.8Ghz. Maybe a little stronger. I'm currently running single-channel RAM w/ one 2GB stick (should be getting another next week or week after) and this board doesn't like overclocking w/ it like this.

The slowdown is basically the hitching. I didn't mean like what FEAR does where it goes from 90 to 9 fps and stays there, but it will drop down for a second when it hitches. I really hope they can fix that in the patch.

I think they're supposed to fix it, along with a # of other things. Hopefully the online play, for one.

[Xion X2]

turn the blower on to 70% that way it will keep things around 50C (at least for me).
Also, 1.20V is hardly scratching the surface here. It's more like 1.238V to 1.250V at 900 GPU.
With Mem at 880 and higher I found increasing the MVDDC one notch above 2.20V helps increase the min frame rates. Not sure about MVDDQ yet but from what I do know increase it a notch helps bring some balance to the MVDDC. Once things are setup rung Unigine at 1680x1050 4x/16x with everything enabled.
Man, that is one loud fan. Even at 50% it sounds like a hair-dryer on steroids. Glad I'm on water.

It didn't make much of a difference, either.. probably because I have very little airflow in my case. I'm just going to have to invest in some ramsinks to keep the mosfets/RAM temps down and maybe a silent 120mm on the side panel to intake a little air on it.

[Bman212121]

At 1680x1050?



Try using the hood cam if you can. Mine stays right around 55fps. I also still have my Conroe clocked at stock speeds (2.4GHz) for right now until I can customize a new BIOS for this board. I'm currently running single-channel RAM w/ one 2GB stick (should be getting another next week or week after) and it doesn't like overclocking w/ it.



I think they're supposed to fix it, along with a # of other things. Hopefully the online play, for one.

Yes at 1680 x 1050 with 4x AA, 16x AF HDR on.

[Xion X2]

Yes at 1680 x 1050 with 4x AA, 16x AF HDR on.
Cool. They're pretty comparable then.

[slaWter]

Try using the hood cam if you can. Mine stays right around 55fps. I also still have my Conroe clocked at stock speeds (2.4GHz) for right now until I can customize a new BIOS for this board, so it should be fairly comparable to your FX @ 2.8Ghz. Maybe a little stronger. I'm currently running single-channel RAM w/ one 2GB stick (should be getting another next week or week after) and this board doesn't like overclocking w/ it like this.

Why do you have to customize a BIOS on that well known overclocker board?

[Xion X2]

Why do you have to customize a BIOS on that well known overclocker board?
This is the Bad Axe 1; it's different from the revision 2. Intel locked certain settings in the BIOS (like voltage adjustment) that can't be unlocked w/o using their development kit and customizing your own.

It's not that difficult.. I've spoken w/ guys who have had success w/ it. I just can't overclock in single-channel RAM mode which is why I've held off. I'm picking up another 2GB G.Skill stick to pair so I'll have 4GB probably some time next week.

[slaWter]

Ah I see. But here comes the obvious question.. Why didn't buy the Rev 2 board?

[Obie]

Cool. They're pretty comparable then.
Are you still getting 55FPS on Sanctuary at 1680x1050? I am getting no were near that. The best I get is 50 FPS. What do you have PCIe frequency set at?

[Xion X2]

Ah I see. But here comes the obvious question.. Why didn't buy the Rev 2 board?

I got this one 110$ off the retail price of $199 brand new off some guy. At the time, I didn't know the BIOS was locked, either.

[Xion X2]

Are you still getting 55FPS on Sanctuary at 1680x1050? I am getting no were near that. The best I get is 50 FPS. What do you have PCIe frequency set at?
PCI-E is set to default.

[Bman212121]

Cool. They're pretty comparable then.

Yea, it's not that far off. I'm probably getting about 5 - 10 more fps, but this card also cost another $200. Either way, if you can vsync @ 60 it really doesn't matter anyway. It will be interesting to see if once you get your C2D OC'ed if you come right up into the 55 - 60 range as well.

[Blacklash]

In the Sanctuary Demo CPU has impacted my score @ 1024x and not @ 1680x. I get the same stock and overclocked when benching 1680x. Overclocking my CPU adds a whole 3FPS to the 1024x bench:p I get 128 instead of 124.9.

My 1680x score is 68FPS. I've posted them before and will again I 'spose.

CPU stock GTX 652|2054 @ 1024x with all on 4xAA|16xAF = 124.9FPS

http://m1.freeshare.us/149fs910881_th.jpg (http://m1.freeshare.us/view/?149fs910881.jpg)

Official results uploaded from the program folder:
http://members.cox.net/malficar3/unigine_20070607_0014.html

CPU stock GTX 652|2054 @ 1680x with all on 4xAA|16xAF = 68FPS

http://m1.freeshare.us/149fs91096_th.jpg (http://m1.freeshare.us/view/?149fs91096.jpg)

Official results uploaded from the program folder:
http://members.cox.net/malficar3/unigine_20070607_0031.html

Fireballs were off.

[Xion X2]

68fps is pretty strong for that. I'm going to bench it again w/ my current overclock @ 900|1950 and see how much of an increase I get.

[XMAN52373]

In the Sanctuary Demo CPU has impacted my score @ 1024x and not @ 1680x. I get the same stock and overclocked when benching 1680x. Overclocking my CPU adds a whole 3FPS to the 1024x bench:p I get 128 instead of 124.9.

My 1680x score is 68FPS. I've posted them before and will again I 'spose.

CPU stock GTX 652|2054 @ 1024x with all on 4xAA|16xAF = 124.9FPS

http://m1.freeshare.us/149fs910881_th.jpg (http://m1.freeshare.us/view/?149fs910881.jpg)

CPU stock GTX 652|2054 @ 1680x with all on 4xAA|16xAF = 68FPS

http://m1.freeshare.us/149fs91096_th.jpg (http://m1.freeshare.us/view/?149fs91096.jpg)

Fireballs were off.

Where to get the Sanctuary Demo from?

[Blacklash]

Where to get the Sanctuary Demo from?

The demo was posted in the below thread:

http://www.nvnews.net/vbulletin/showthread.php?t=91925

[XMAN52373]

Ok, My D3D9 run was 34.8, OGL was 35.6 same settings as BlackLash on a X2 3800+ @ 2.5, BFG8800GTSOC 640MB, 2GB DDR500 and XPPro/SP2.

PS> How did you upload the HTML file or did you save it as a jpg?

[Blacklash]

Ok, My D3D9 run was 34.8, OGL was 35.6 same settings as BlackLash on a X2 3800+ @ 2.5, BFG8800GTSOC 640MB, 2GB DDR500 and XPPro/SP2.

PS> How did you upload the HTML file or did you save it as a jpg?

It's not .html, just a capture of what I got when I opened the bench results. So it's a jpg. I will provide the official .html from the program folder by uploading it to my personal webspace.

Below are the official .html results from the benchmark:

Bench 1
http://members.cox.net/malficar3/unigine_20070607_0014.html

Bench 2
http://members.cox.net/malficar3/unigine_20070607_0031.html

[XMAN52373]

It's not .html, just a capture of what I got when I opened the bench results. So it's a jpg. I will provide the official .html from the program folder by uploading it to my personal webspace.

Below are the official .html results from the benchmark:

Bench 1
http://members.cox.net/malficar3/unigine_20070607_0014.html

Bench 2
http://members.cox.net/malficar3/unigine_20070607_0031.html

Got almost a 2fps increase upping my core clock 30Mhz. Now it runs at 575/1350 in 3d mode, 470/1134 in 2d mode(that would be the rop/shader clocks. Memory is at 864. Man I need to try water cooling.

[Blacklash]

Got almost a 2fps increase upping my core clock 30Mhz. Now it runs at 575/1350 in 3d mode, 470/1134 in 2d mode(that would be the rop/shader clocks. Memory is at 864. Man I need to try water cooling.

Clocks may not make up for missing shaders in certain benches. A guy in the thread I linked earlier with the bench download was getting 82FPS @ 1024x with a GTS @ 630 on the core and I believe a guy with a stock GTX was getting 113 or more @ the same res.

[fivefeet8]

Sanctuary Demo. Settings Shown in pic.

AMD x2 3800+@2.4ghertz, 8800gtx@625/2022, Vista 32bit.
http://img395.imageshack.us/img395/6286/sdemobn4.th.jpg (http://img395.imageshack.us/my.php?image=sdemobn4.jpg)

With the 8800gtx@stock I get about 55 fps.