8800GT Official Reviews

[mascaras]

yep tnks

Temps @ fan Auto @ 2minuts atitool scan for artifacts

[Pacha]

I guess following the atitool fan override table you are not at 100% fan speed, nice temps indeed.

it just confirms that the chip runs quite cool, good point.

[Soulburner]

I don't agree. Mine will run up to 75c at 50-60% fan speed. Any higher and its a hairdryer - way too loud.

[trans am]

Yeah. Use riva 2.06 to force 100% fans and oc core/ shader and memclocks and once you set them open ati tool and check for artifacts.

http://www.guru3d.com/index.php?page=rivatuner

[mascaras]

I guess following the atitool fan override table you are not at 100% fan speed, nice temps indeed.

it just confirms that the chip runs quite cool, good point.

im with auto fan @ 30%

[GAR]

at 756/1836/2000 ati tool artifacts, but crysis plays for an hour, no artifacts nothing, im usising vista x64 is ati tool a reliable nvidia tool???? i mean it is ATI tool???? what do you guys think, i need some expert advice.

[Plywood99]

Don't know if it has been posted yet or not, but a rumor is going around about a bad batch of cards getting into retail. Rumor states that 2d is fine but can be slow, but going into 3d bsod's the computer.

This is exactly how my XFX card acted. Bummer...


Ply

[RPGWiZaRD]

at 756/1836/2000 ati tool artifacts, but crysis plays for an hour, no artifacts nothing, im usising vista x64 is ati tool a reliable nvidia tool???? i mean it is ATI tool???? what do you guys think, i need some expert advice.

ATITool has always been the utility I trust most as it's the most stressful for gfx card I was able to find, like you say it's even better finding artifacts than playing Crysis, if it's atitool stable it's stable for all games too. Like my 7900GTO would be artifact free in 3DMark at 700~705MHz core setting but require 690 to be Atitool artifact free. Great utility for artifact scanning and temp measurement. And a good thing you don't need to run it for long to know if the clock is unstable or not, even 15mins is a very good indication, testing over 1hr to me seems totally unnecessary in this utility.

[Eagleclaw]

I have had one time while testing my 8800GT clocks last week where atitool passed 30 mins but Crysis would hard lock after 10 mins of actual game play until I lowered my clocks.

So now for me I am using both to test stability.

[RPGWiZaRD]

There's still a possibility there's a driver/game incompability too. Maybe maybe not, but you can't deny that either.

[adamsleath]

at 756/1836/2000 ati tool artifacts, but crysis plays for an hour, no artifacts nothing, im usising vista x64 is ati tool a reliable nvidia tool???? i mean it is ATI tool???? what do you guys think, i need some expert advice.

im opposite as i can have artifact free atitool and yet the occasional ingame glitch...eg.
720/1800/1950 in tf2..will play fine for say 15-30 minutes and then ill get a fatal glitch...

but maybe with your oc right on the edge of stable it's different.

does orthos stress gcards??

what other gcard specific test is there?

anyway if it's stable in the games you want to play who cares? the real stress test is the actual app...

theteamaqua says "crysis stable" - i think it is an apt term.

[Pacha]

at 756/1836/2000 ati tool artifacts, but crysis plays for an hour, no artifacts nothing, im usising vista x64 is ati tool a reliable nvidia tool???? i mean it is ATI tool???? what do you guys think, i need some expert advice.

why bothering with atitool for artifact scanning for hours? it just makes the gpu generate maximum heat, I don't find it much more interesting than that.

as you'll certainly get sligthly higher clocks in 3dmark than your games, use your games to test stability. just alt-tab to increase the clocks and go back in your game to test, until it fails. stability in the games you play is what you are looking for, don't you?

[GAR]

anyone, replace the paste on the card???

[RPGWiZaRD]

Well ATITool has for me been very accurate for the last 3 cards I've been using and found artifacts a bit before any game/benchmark I've tested so for me it's been the optimal test. Well perhaps it's different with Geforce8 series as ATITool perhaps might not stress the shader domain that heavily so that could be the reason for you that Crysis would show glitches earlier due to unstable shader clock. That's my best guess, using both sounds like a good idea.

[adamsleath]

i thinned out the paste on the core attempting to make better contact using stock hsink. i thought about putting on diffrent goo but i was too lazy

i could see by the impression the core had made in the paste that the core was not sitting flat onto the heatsink as it came to me out of the box.

[Pacha]

always, since mounting my 1A-SL2 every time. as5 each time.

[Eagleclaw]

All good points. Atitool IMHO is a very good stress tool and I have used it for a long time and I will continue to use it along with Crysis for my tests.

[Pacha]

Well ATITool has for me been very accurate for the last 3 cards I've been using and found artifacts a bit before any game/benchmark I've tested so for me it's been the optimal test. Well perhaps it's different with Geforce8 series as ATITool perhaps might not stress the shader domain that heavily so that could be the reason for you that Crysis would show glitches earlier due to unstable shader clock. That's my best guess, using both sounds like a good idea.

I've already experienced the reverse situation of adam, ie card stable in atitool's artifact scanning and not in games. with a X800 Pro.
this was my first try of this in fact. coupled to my first and last test of "find max gpu" which had increased the gpu freq so much that I didn't have any artifacts, just lines all over the screen like sort of a test card...


orthos or occtpt are ok for me as they test the cpu cores at their maximum and are reliable enough to ensure all others apps will run fine, but gpu talking, I'm not sure that many other games will stress your gpu more than crysis does atm, so playing your current "heaviest gpu duty" game should be enough in my opinion

[mascaras]

crysis
GPU benchmrk
8800GT default (625/900)
1680x1050
game settings high
AA & AF OFF
Q6600 @ 3.6ghz
Forceware 169.04
windows XP


MIn:19
average FPS 29.66
Max 34.39






another proof of 169.04 are bug , (look ate bushes and car rearview )


NFS Pro
Q6600@3.6ghz
forceware 169.04
CAT 7.10
HD2900 Dfeault
8800GT OC Default
windows XP




8800GT




HD2900XT



regards

[GAR]

@ mascares, whats your max OC??? mine seems to be 720/1782/2000

[mascaras]

Crysis
Q6600@3.6ghz
forceware 169.04
CAT 7.10
HD2900 Dfeault
8800GT OC Default
windows XP
1680x1050
Game setings : High
AA & AF : OFF


8800GT







hd2900xt

@ mascares, whats your max OC??? mine seems to be 720/1782/2000

didnt test yet , next thing



edit: looks like we have a good OC´r



only core Test :



FAN @ 90% --- 725mhz core -- 1750mhz shader - 1800mhz memory --- FULL Temp = 54º









FAN @ 90% --- 740mhz core -- 1750mhz shader - 1800mhz memory --- FULL Temp = 54º














750mhz core >>> artifacts just a litle bit after 30 seconds @ atitool

so i would say max stable clocks without artifacts @ Stokcooler (90%) = around 745mhz core

[Diverge]

um, those aren't temps that really represent the stock cooler. let it run for 10+ minutes until your temps stabilize and the graph is flat. also, the 3d (fuzzy spinning cube) generates more heat then artifact scanning ;p

also, just cause it doesn't generate artifacts in ati tool doesn't mean it is stable in game. I can clock my card over 770mhz, run 3dmark06, ati tool, but it will crash in crysis or hellgate london.

[mascaras]

also, the 3d (fuzzy spinning cube) generates more heat then artifact scanning ;p
.

hmmm, i think scan for artifacts more heat generated then 3d View ...

[theteamaqua]

47C max during crysis, for like 15 min

http://i121.photobucket.com/albums/o...Shot0042-1.jpg

it then dropped to 45/46C in some areas, and this is with HR 03 regular only .. cant wait for that 6 heatpipe GT... i have VF900 on my 2nd card, and its actually like 9C hotter than my 1st card in 3dmark06, lostplanet ....

[GAR]

47C max during crysis, for like 10 min

http://i121.photobucket.com/albums/o...Shot0042-1.jpg

it then dropped to 45/46C in some areas, and this is with HR 03 regular only .. cant wait for that 6 heatpipe GT... i have VF900 on my 2nd card, and its actually like 9C hotter than my 1st card in 3dmark06, lostplanet ....

Hey did you get a better max OC with the HR03, or was it the same as stock???