• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

5970 vs 5870Xfire vs GTX460SLI. Results starting to come in!!

Page 6 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
lol, no Ares, just two EAH5870's V2. I will set down and run both ways, with AA and without. I just forgot to turn AF back on when I reinstalled the drivers.
 
james1701 said:
lol, no Ares, just two EAH5870's V2. I will set down and run both ways, with AA and without. I just forgot to turn AF back on when I reinstalled the drivers.
Click to expand...

Yeah, just take your time, nobody is rushing you at all. Take a few runs in each bench with different settings. Start out Ultra Maxxed on everything and back off if you need to on certain things. When you have as much eye candy as you can and the game still remains playable to you, that's when you get a Fraps capture going. Some games you might run so easily that you can't turn anything else up, you've done it all.
 
resent. without AA and with 4x AA. AF is 16 on all games. I did not use physx on Metro. I forgot to include BF BC2, hold on, and I will get it.

Edit, BC2 in second Email.

CPU 980X 4ghz, Memory 6gb 2000mhz, crossfire Asus EAH5870 V2.
 
Last edited:
hey I will run benches as soon as my 2nd 5870 arrives. Also, dont mean to derail the thread or anything, but how did yall get to 1000 core on the 5850s? How much vcore was needed? Is it generally easier than 1000 core on a 5870? Currently at 950 core on mine, but not sure how much farther I can go.

Also, is there a recomended max to memory frequency on the 5870? I heard that beyond 1300 there is no point.

Thanks, and hope to be benching soon!
 
Shmee said:
hey I will run benches as soon as my 2nd 5870 arrives. Also, dont mean to derail the thread or anything, but how did yall get to 1000 core on the 5850s? How much vcore was needed? Is it generally easier than 1000 core on a 5870? Currently at 950 core on mine, but not sure how much farther I can go.

Also, is there a recomended max to memory frequency on the 5870? I heard that beyond 1300 there is no point.

Thanks, and hope to be benching soon!
Click to expand...

Personally, juice, juice and more juice 😱

I need 1.35 to keep it happy at 1Ghz core (well, laziness means I haven't exactly played around with it, but slapping the slider right over stopped the crashing and that was enough for me!). There's no way I'd use that for regular gaming given the VRM temps under sustained load (maybe with water on the VRMs 😉). I believe power draw goes out the window (my 500W Antec Earthwatts certainly spools up like I have never heard before!) with that kind of voltage/core speed combo 😱

Happy to do so for benchmark purposes in the name of scaling and science for the Video forum, of course 😉

For everyday use and sane VRMtemps/voltage I usually go with 930 or so.

I have not had time nor inclination to investigate the upper limit on my RAM, 1225 seemed to work fine and I thought I read somewhere that much over 1200 was blowing in the wind...as I understand it sometimes 5850s struggle to hit those kind of speeds as the RAM voltage is 'hardset' lower than the 5870, but maybe because mine is an XFX Black Edition with stock 765/1125 clocks that's different? Who knows 🙂

Your clocks sounds excellent mate, relax basically 🙂
 
Last edited:
ya, 1250 is actually the stock memory on the card. I believe I got the black edition, though I ordered the xxx edition and got that box. 900 was stock core 😀

Soon I get to try a visiontek card, once it arrives. Who knows, might be a great ocer😀
Reference cards are very hard to find these days.
 
Sorry guys, been very busy last few days. I will start adding the additional benches this afternoon.

Keep em coming!!

Keys
 
Just benchmarked Mafia II with my standard settings. Received 127 FPS. Settings are 1920x1080, PhysX off, AA off, AF 8x, everything else at max. Specs in sig.
 
I think we should be trying to bench at the exact same settings, no ?

Adam, can you bench the same settings that i did so we can compare easier ?
 
Well, the sites accurancy isn't all that great then. Shows them all at "19x12" for some odd reason.

I reran the test with 16x AF and AA on. Everything else the same. Received 122.6 FPS.
 
I will continue to fly the flag and bench 19x12, that's real man's resolution 😉

Unless people want me to do 19x10 as well of course 🙂
 
Metro 2033

DX 11- Tessalation- DOF- AF4 - AAA

1680x1050- Frames, Time (ms), Min, Max, Avg
13757, 315376, 22, 71, 43.621

1920x1200- Frames, Time (ms), Min, Max, Avg
11103, 313709, 16, 57, 35.393

2560x1600- could not run..... too choppy i think i ran out of vram.. nevermind, got a run in you just have to wait like a minute for the textures load before you run the game. however i left on AF16 so the run is in that section



Dx11- Tessalation- No DOF- AF4- Very High- AAA

1680x1050- Frames, Time (ms), Min, Max, Avg
19630, 321331, 26, 144, 61.090

1920x1200- Frames, Time (ms), Min, Max, Avg
17649, 330551, 20, 118, 53.393

2560x1600- Frames, Time (ms), Min, Max, Avg
11444, 329377, 10, 74, 34.744


DX11- Tessalation- No DOF-AF16-Very High- AAA

1680x1050- Frames, Time (ms), Min, Max, Avg
18622, 316029, 25, 127, 58.925

1920x1200- Frames, Time (ms), Min, Max, Avg
15660, 313392, 20, 104, 49.969

2560x1600- Frames, Time (ms), Min, Max, Avg
10678, 315747, 10, 73, 33.818

Now a matching run to Annisman*

1920x1200- DX11- DOF- Tessalation- Very High- AAA- 16AF
Frames, Time (ms), Min, Max, Avg
11221, 320120, 15, 58, 35.052

2560x1600- DX11- DOF- Tessalation- Very high- AAA- 16AF
Frames, Time (ms), Min, Max, Avg
5488, 329194, 2, 29, 16.671
 
Last edited:
after toying around i realized how inconsistant benching 2033 can be in the area we chose.

depending on where the player decides to stand in the first area when the monsters attack(lots of volumetric lighting) the fps can be drastically different. I stand on the top of the stairs staring straight at the monster on the second level.

If i stand under the second level my fps are 20+ fps higher. I could artificial inflate my score like this by lingering in some areas and avoiding others.

During my testing I tried to tax my setup as much as possible and make the results as consistant as possible too.

I think we need to find a better area to bench this game.
 
I suspect we have overloaded poor keys with our benchmarks, kudos to you keys but it's a huge collection of data!
 
What I'm really hoping for, is that everyone agree on a standard. Same test, same page, same settings wherever possible. All these results I'm adding kind of feel like nothing is getting done in the way we'd like, which to me, is final, across the board comparable results. So, I was hoping you guys would have a pow-wow and work out the details.
 
I still think we should all be benching 1920x1200 (or 1080 if possible)

4XAA, 16AF in every title, no exceptions, and highest settings possible. Do that for every game, then decide on where to bench in the game. As you can see by my entries, I try and specifically state where and how in a game I decide to bench it.

Physx off of course, because ATI cards aren't allowed.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top