Andre, try some of the low cards, right now and do everything you are doing with the high cards, see if anything changes.
Printable View
Andre, try some of the low cards, right now and do everything you are doing with the high cards, see if anything changes.
I already have tried low card and high card on air for 3D03 3D05 3D06.
http://www.xtremesystems.org/forums/...&postcount=397
Only 03 is big different.
imagine you go buy 20 cards and none of them do bug high EEEK hahahah
Started air testing my cards individually again last night.
Rampage Extreme
E8600 @ 4000MHz
DDR3 @ 7-6-6-15
Gigabyte 4870X2 @ 800/1000 TOP BIOS with stock copper heatsinks but no mounting plates or fan/shroud
PCI-E slot #3
XP SP2 no updates
Catalyst 8.9 driver
ATT high performance settings
Card #1 horizontal HS fins
Test #1 - 93k
Test #2 - 93k
Test #3 - Hang on trolls
Test #4 - 93k
Card #2 - vertical HS fins
Test #1 - Hang on proxy
Test #2 - Hang on proxy
Test #3 - Hang on proxy
Will remount heatsinks tonight in same orientation as card #1
Once i find out the results of card #2, i'll x-fire them to see if still low or recovered back to high.
didnt realise you can change stock heatsink direction
I took photos for an experiment write-up here, but havent hosted it yet.
Here's what i mean though from an old photo i took but with one heatsink rotated 180 degrees from the other.
The "good" card from last night was running with both orientated as below (horizontal fins) but not 180 degrees out.
http://img.photobucket.com/albums/v310/T_M/IMG_6532.jpg
did you have any heatsinks on all the other areas eeek
Nope, not required for low volts/clocks
lol just broke 100k with the "bad" card (seems core 2 likes more volts than the other) at only 860/1010 and 4425MHz on CPU
That sounds like a "good" score to me. :)
Yep both my cards are running high individually and consistently.
Both get exactly the same scores and top out at 860 core (870 causes crash to desktop and resolution change, or low score and resolution change).
Next up with xfire retesting.
How's that look?
All stock on a Sapphire X2. 750/900, 8.9.
Doesnt look good mate, but try with a bit higher CPU.
At 780-1000 and 4GHz i get 91k easily.
Would if I could, cant until I get my other ram in.
Also this is a dual socket system.
COMFIRMED.... :(
"GOOD VGAs" and "BAD VGAs"....
Yeap....gprhellas brought me his VGA today which was doing ~ 108000 SINGLE...
At 3600MHz and 850/1000
Mine (5 of them :( )
86361
1067.5
747.4
570.8
621.1
gprhellas'
93285
1067.0
859.2
630
621.0
Bad news.....
I didn't toughed ANYTHING.....
I just pluged my VGA boot into windows and becnhed 2k3...
Shuted down and pluged gprhellas' VGA...
Boot into windows and benched 2k3....
THAT simple.... :(
So to clarify is it
A - There are "good" cards and there are "bad" cards
B - All cards start off the same but some become degraded
C - To be honest John, we really don't have a clue
As for me, the "A".... :)
IF you are LUCKY, you get a "GOOD VGA" .... Chances though to find one are about 10% - 5% MAX.... :(
Maybe it has something to do with the week of production of the PLX chip as Mickeymouse have told.... :(
SO ALL "high scores" are LEGIT on hwbot (exept the bugged ones)..... :)
.
Why do I forsee a "batch database" for ATi cards emerging we'll soon be trying to figure out good weeks for cards now too :shakes:
yes, that was also reported and discussed on ocx regarding the VGAs
Andre got some sweet VGAs
Im not so convinced.
Both mine get high runs on single and xfire.
Then i put them on cold and got low.
Now back on air they're high again.
Just ran xfire 860/1020 and 4400MHz = 132712
Me neither T_M, stil high and low scores with same card on air.