Tim, your situation really did baffle me...I thought for sure you'd have been way past 160k once you got going judging from your air scores.
But of course, my scores remain "high" cold, so the theories of heat causing issues is out the door imho.
The next question we need to ask is does this only affect 03? I asked both Vince and George if they had tried Vantage xtreme yet...in performance there is no difference, but perhaps in high and xtreme there might be??
Probably not, Gautam, there's no difference in 01/05/06/AM3.
Where courage, motivation and ignorance meet, a persistent idiot awakens.
Not even in single card I take it.
In which case this is really weird, and something that I couldn't see ATi giving a damn about.
An obscure issue in one, single, five year-old benchmark.
I think I got a good one by the looks of your scores at 3600, just masked by my FB-DIMM powered system holding it back a bit. YayOriginally Posted by hipro5
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....
All along the watchtower the watchmen watch the eternal return.
Today a friend of mine bought a new 4870x2 card . An Asus one to be more specific.
So ... We plugged the card and tried 3dmark03 . First run at 800Mhz core gave us about 740fps at GT2 . I started o/cing the card and at about 850 it crashed .
I turned the PSU off ... and then on again ... booted in windows and run again GT2 at 800Mhz core . Result ??? About 850FPS GT2.
I shut the computer down again . PSU off ... and then on . Run again ... and GT2 at 740FPS.
Again the same story . Shut down ... power off and power on . Run again and 740FPS GT2.
And again I power down the PSU ... then on again ... and this time we got 850FPS !
Powering down the PSU is the key for fast 3dmark03 runs in our case .
Not necessarily, could be rebooting helping?
Hmmm ... Didn't try that , but will do .
Just to confuse you even more Gautam, yesterday i ran some single card LN2.
Gigabyte 4870X2 voltmodded and Asus TOP bios set to 860/1020.
2 x Tek-9 4.0's direct mounted and card insulated with vaseline/laquer.
QX9770 and Rampage Extreme.
Started off using '06 and did a run at 5400MHz, and (from memory) about 900/1020 and scored what i had expected (29.7k). I think the card was still cooling down while running so probably around -50 to -60C.
Once the cards were down to -90C (-100C+ caused no boot) i did a run at 940/1020 and for some reason my score dropped to 29.6k (CPU score was still same, just lower GTs)
After a lot more playing i found that if i let the card warm up to -50C and sit between that and -60C, my score boosted back up again and i got a run at 980/1020 giving 29.8k http://www.hwbot.org/result.do?resultId=781580.
To test my theory i ran some '03 quickly.
If i ran the card back down at -80C+ my GT2 would suffer badly (900fsp to 700fps start).
If i ran the card up at -60C, GT2 would boost back up and run 900fsp to 1100fps start, but i cannot conclude anything cos i also fiddled with voltages a bit to avoid crashes.
I did not complete an '03 run due to white screens at GT2 end.
Next up i went on to Vantage and ran calico successfully at -60C (-50C caused freeze at explosion)
http://www.hwbot.org/result.do?resultId=781585
Right at the end when i ran out of LN2 i reran some '03 and was bugged low even though temps were -50C????
Last edited by T_M; 10-12-2008 at 07:56 PM.
cold bug past -60C maybe?
Could be temperature - could be voltage
Unfortunately i wasnt very systematic
Heh, heh, heh...... I sold ALL of my 4870x2s and then went, tested and bought 2 x "good" Gigabytes 4870x2....
I tested the first one and no go......the second one was good, the third one was good and the fourth was not good......So I had a 50% luck yesterday night....
GOOD thing is that I FIRSTLY tested them and THEN I bought them.....
Now I'm searching fro two "good" ASUS TOP ones....
Last edited by hipro5; 10-14-2008 at 05:45 AM.
Well, at least there is some kind of a solution if your cards bug low: buy another one.
It's a pretty crappy solution, though
Where courage, motivation and ignorance meet, a persistent idiot awakens.
Yes....
My chances were ~25% ...I think I was lucky....
INTEL PWA FOR EVER
Dr. Who my arss...
.........
I have 1 bad and one good.., assembled a puter for a friend last week.., same there.., one good.., one bad...
Called a friend who also have a pair.., one good and one bad..., different brands all together..
How high are odds for this "one bad", one good"...? :-))
so did you swap the good and bad with that friend
he prolly doesnt care
Ha.., no.., I didnt.., he got so suspicious when I tried to convince him that he had absolutely no need for this "good" card.., haha...
But.., I will attack my other friend from a different angel..
Ha.., no.., I didnt.., he got so suspicious when I tried to convince him that he had absolutely no need for this "good" card.., haha...
But.., I will attack my other friend from a different angel..
jackass
why dont you just show him this thread
hmmm maybe not
How do you tell if you have good ones or bad ones? I done 136,641 @ 4545mhz quad core quadfire 778/950 http://service.futuremark.com/compare?2k3=5959232
He, he, he............
INTEL PWA FOR EVER
Dr. Who my arss...
.........
Looks like I'm suffering of low GT2 & GT3 scores with my system:
E8600 @ 5871 MHz (9,5 x 618 MHz)
Rampage Extreme (Performace Level 9)
Corsair DDR3 (7-6-6-18 1T) 927 MHz
Asus Radeon HD 4870 X2 TOP (855/995 MHz)
96218 marks:
http://www.hwbot.org/result.do?resultId=782513
I've tried in XP & Vista, PCIe clocks 100 - 120 MHz, etc but no help..
Looks like I'm missing +200 FPS in GT2 and +100 FPS in GT3 compared to some other people in hwbot.
I have two other 4870 X2 cards and I'll try them next (Sapphire 4870 X2 and another Asus 4870 X2 TOP)
Last edited by Sampsa; 10-15-2008 at 11:08 AM.
Favourite game: 3DMark
Work: Muropaketti.com - Finnish hardware site
Views and opinions about IT industry: Twitter: sampsa_kurri
My 05 single card score seems to be ok: http://www.hwbot.org/result.do?resultId=782514
Last edited by Sampsa; 10-15-2008 at 11:08 AM.
Sampsa, just to check can you take your "bad card" and remove the cooler, then just reattach the copper GPU heatsinks?
Blow a fan across both heatsink while you run and see if you get better scores.
Nothing change that....... I've done it already....
I've even put 2 x other heatsinks on both cores and LEAVE the PLX chip WITHOUT a heatsink so to get realy HOT.......NO GO.....Same flucken different day.......
ONLY other VGAs and that's it.......Period....
P.S. EVEN is someone has a "good" VGA and a "bad" VGA, IF he wants to run them in CFX, he'll get the score of TWO "bad" VGAs.....
INTEL PWA FOR EVER
Dr. Who my arss...
.........
Ok, I tested all my three 4870 X2 today (one default Asus TOP, one volt/bios modified Asus TOP and one default Sapphire 4870 X2).
Looks like my failure rate is 100 %
I tested each card with E8600 @ 3,6 GHz (6 x 600 MHz), Rampage Extreme, 4870 X2 @ 790/915, Win XP + SP3. All cards performed similar.. ~710 FPS in GT2. I was so desperate I even tried PSU shut down trick but no help..
edit: looks like my results are aligned with hipro's bad card result @ 850/1000 clocks.. so no 03 for me
Last edited by Sampsa; 10-16-2008 at 02:37 AM.
Favourite game: 3DMark
Work: Muropaketti.com - Finnish hardware site
Views and opinions about IT industry: Twitter: sampsa_kurri
Posting Permissions
Reply With Quote
Bookmarks