Request for 6850 xfire benchmarks.

[erf]

I was kinda hoping someone with xfired 6850's might be bored enough to run some benchmarks.

I'm shopping for my next GPU solution and am pretty torn. If I wasn't held back by chipset/motherboard I'd probably be going with gtx460's in SLI but as far as I can tell my motherboard doesn't support sli unless I use a hack.

That has brought 6850's into the equation. The only review/benchmarks I could find concerning xfired 6850's that included a record of MINIMUM frames per second was one obscure Aliens vs. Predator benchmark on Tomshardware.

This benchmark showed that 6850's in crossfire although having a higher max framerate, suffered horribly when it comes to minimums. The minimum framerate on xfired 6850's dropped 2 times as much as the single card solution. This was not a problem with the sli'd 460gtx's.

I'm all about some high frame rates but if I'm having to sacrifice for a lower minimum frame rate I'm thinking I'll either pass on xfired 6850's and get gtx 460's or go with a single 6870. I really want to run a dual gpu setup and the 6850's would be perfect for my situation but I'm not gonna dump that much money into something that's not practical.

I'm hoping that the one benchmark ran by toms may not be the norm. Anyone care to contribute???

Thanks in advance.

 

[Flyinfinni]

I'll run some when I get my second card next week.

 

[erf]

I'll run some when I get my second card next week.

Is the minimum frame rate thing something you just didn't consider or do you know something I don't???

 

[Flyinfinni]

I know that with crossfired 5750s, it worked great and scaled at ~80%, which is great. I played with that for more than 6 months and never had a problem. I didn't record minimum frame rates with it, but honestly, everything ran better with 2 cards than 1. If there was a place where frames dropped to un-playable on xfire, they were still unplayable on a single card. I'll see if I can dig up some old benchmarks from those cards and see if I have min numbers on them.

EDIT:
From old Dirt 2 benchmarks (granted- canned bench from the game, not actual gameplay, sorry) min frames from a single 5750 was 27.7, average was 34.7. Crossfire 5750s min was 47.3, Ave was 61.3. It definitely felt smoother all around with crossfire. As a reference, my 6850 (OCed a bit- probably at 850/1200) was 51.4 min and 64.7 ave.

 

[Valset]

http://www.hardwarecanucks.com/foru.../37286-amd-radeon-hd-6870-hd-6850-review.html will give you them on a single card.

also I posted this in your other thread. keep it in mind when your looking at the numbers. Minimum framerates are only useful if the site that mesures them does it right

your making way too much of that. all dual card setups do this and it is almost never measured correctly in canned benchmarks. and the 16fps isn't any more playable then the 6fps. usually this happens at some point or other in the game (save points, load points, map changes and such) look at some of the H reviews that do real world playability http://www.hardocp.com/article/2010/..._card_review/4 all three single cards do this badly here. so when you see the "minimum frame rates" you need to be well assured how they are measuring that. if they are looking at actually slowdown in gameplay or something else entirely. This is why you should always wait for the H review .

BTW this actually works in reverse as well. When the 3870X2 came out it had good benchmarks in crysis. (kicked a lot of ass). the problem is that when you tried to play it did suffer some minimum frame rate issues. They issues simply did not show up in the benchmarks. read here http://www.hardocp.com/article/2008/...g_benchmarks/6

having said all that I agree that a single 6870 is probably a better choice along with a monitor upgrade.

 

[erf]

http://www.hardwarecanucks.com/foru.../37286-amd-radeon-hd-6870-hd-6850-review.html will give you them on a single card.

also I posted this in your other thread. keep it in mind when your looking at the numbers. Minimum framerates are only useful if the site that mesures them does it right

your making way too much of that. all dual card setups do this and it is almost never measured correctly in canned benchmarks. and the 16fps isn't any more playable then the 6fps. usually this happens at some point or other in the game (save points, load points, map changes and such) look at some of the H reviews that do real world playability http://www.hardocp.com/article/2010/..._card_review/4 all three single cards do this badly here. so when you see the "minimum frame rates" you need to be well assured how they are measuring that. if they are looking at actually slowdown in gameplay or something else entirely. This is why you should always wait for the H review .

BTW this actually works in reverse as well. When the 3870X2 came out it had good benchmarks in crysis. (kicked a lot of ass). the problem is that when you tried to play it did suffer some minimum frame rate issues. They issues simply did not show up in the benchmarks. read here http://www.hardocp.com/article/2008/...g_benchmarks/6

having said all that I agree that a single 6870 is probably a better choice along with a monitor upgrade.

Hey, thanks for the responses. I have another question for you. How do you like that 28" HannsG???

I just found that thing for $219 at Mwave.com.

http://www.mwave.com/mwave/skusearc...-281DPB / HG-281DJB LCD 28W HDMI W HDCP DVI S

It's refurbished, but, in this case I could go for that.

edit: sorry just noticed that your monitor is a Hannspree....haha.

 

[Valset]

Hey, thanks for the responses. I have another question for you. How do you like that 28" HannsG???

I just found that thing for $219 at Mwave.com.

http://www.mwave.com/mwave/skusearc...-281DPB / HG-281DJB LCD 28W HDMI W HDCP DVI S

It's refurbished, but, in this case I could go for that.

edit: sorry just noticed that your monitor is a Hannspree....haha.

if its the same as the hannspree not so well. loved it for the first couple years but it has developed an issue where the monitor menus randomly show up on screen. I don't think its an issue with the pannel as much as the first gen electronic. dues to crappy support I will not buy another hanspree. but others use this panel to very good screens. if your looking for a large 2.3mp monitor this panel is awesome for a TN. but I would get the INC or the like before I got a Hannspree. I am not sure if HannsG is the same or not. Names are similar

 

[EquaLiZr]

Its been totally worth it for me I play my games at 1920x1200, 4xaa, 16af and rarely dips below 60fps. In most of my games I can now have all eye candy turned on.