Page 1 of 1
An odd question about benchmarking x800s
Posted: Wed Jan 31, 2007 5:47 pm
by dgood
so why is it that they offer x800 master cards but i've almost never seen the benchmarks of x850/x800 crossifre set ups on sites? besides the fact that I know they aren't that great and they use the wonderful dongle I just wondered why I never see them anywhere not even on toms hardwares big vga chart list is there a cf set up of two x800 (1 and a master card) or 2 x850s ( different master card for them).
Posted: Wed Jan 31, 2007 8:45 pm
by cyberneticimplant
I think it might be due to the fact that Crossfire is just not as popular as SLI.
http://www.steampowered.com/status/survey.html
NVIDIA SLI (2 GPUs) 13,664 96.00 %
NVIDIA SLI (3 GPUs) 5 0.04 %
NVIDIA SLI (4 GPUs) 2 0.01 %
ATI Crossfire (2 GPUs) 563 3.96 %
There are a few reviews though. Here is a review of a X850 Crossfire setup.
http://www.hardocp.com/article.html?art=ODE1
Posted: Wed Jan 31, 2007 8:46 pm
by dicecca112
yeah you just don't see as many crossfire reviews on cards less than x18 series with ATI. You see a ton for SLI though
Posted: Wed Jan 31, 2007 9:12 pm
by DMB2000uk
And crossfire before the x1800 series was poorly implemented, with restrictions on refresh rates at certain resolutions, and if i remember correctly a maximum resolution of 1600x1200. To top it all off, I don't think that the performance gain was justifiable to get a master card in most situations.
Dan
Posted: Wed Jan 31, 2007 9:38 pm
by Bwall
Actually I've just been messing around with X850 CrossFire and while I have seen some pretty impressive gains running it over a single card it still has some issues. Look for it in an upcoming article.

Posted: Wed Jan 31, 2007 10:31 pm
by dgood
sweet I was just thinking if I got a cheap cfx3200 lanparty ut on ebay for more overclocking fun, i might buy a master card and dink with it because you can get them for liek 80 dollars or something. mostly just wanted to see the results in a large picture and its just not available since almost nobody ran them.