HELP! 8800GTX GPU Volt mod gone wrong!

[KnownKiller]

HELP!!!!

I just did the volt mods to my Evga 8800GTX card and i have a small tiny prob.. I broke that damn tiny black resistor thats used for the GPU volt mod!

Now i have a few spare card lying around that i can use for parts.. Just find the identical size black resistor? Or could i just use a fixed resistor to replace it?

Oh btw.. Broke the damn thing by tuggin on the soldered wire trying to move another wire that was runing under it. Just tore that resistor right off and it looks like it snapped the end so i dont think i should use it again.

http://resources.vr-zone.com/Shamino/8800gtxvmod/1.jpg

[xilix]

Damn that sucks. If something like this happened to me I'd be running to Viper John real quick and see if he could fix it for me. If you can find that exact resister then you should be able to replace it just fine as long as you can solder that small.

[KnownKiller]

Damn that sucks. If something like this happened to me I'd be running to Viper John real quick and see if he could fix it for me. If you can find that exact resister then you should be able to replace it just fine as long as you can solder that small.

Yeah.. Someone at xtremesystems told me that if the resistor reads 34 ohms then its still fine. Mine does but its still attached to the wire i soldered to it on one side.

Dont know if i should attemp to solder it back on with the wire still attached or take it off. Though its so small i'm prob better off trying it as it is.

I also just had fun trying to find the exact resistor on an old crappy wireless card. It had a ton of these on there and i had to rip them off 1 at a time and test them with a MM.. ofcourse there all of different ratings.

Looks like trying to find one is gona suck.

Thats why i was wondering if its possable to just use a fixed resistor of the same value or something..

Back when i seen someone use a fixed resistor to replace a broken resistor on a 7800GTX when he had probs with the mod.

[KnownKiller]

Hey thanks xilix! I just shot off a PM to viper john to see if he has any ideas.

[Spawn-Inc]

you shouldn't have to rip of the resistor to test it, just put the leads at each end.

if the original resistor has the same reading as what was said (34ohms) then just get a pin or something small and put some flux on the two spots, be careful not to connect the two solder points, and drop the resistor on the board in that spot. if you have more than one soldering iron or if the tip can be removed try to find a copper slug or something similar and shave/grind it to a fine point, like a pin, and try to solder it back on.

practice with your old wireless card if you don't feel you can do it the first time.

[KnownKiller]

you shouldn't have to rip of the resistor to test it, just put the leads at each end.

if the original resistor has the same reading as what was said (34ohms) then just get a pin or something small and put some flux on the two spots, be careful not to connect the two solder points, and drop the resistor on the board in that spot. if you have more than one soldering iron or if the tip can be removed try to find a copper slug or something similar and shave/grind it to a fine point, like a pin, and try to solder it back on.

practice with your old wireless card if you don't feel you can do it the first time.

Thanks for the advice.

Yeah i thought i could'nt get a reading on them untill i took the resistor off.. Prob was i had the DMM set to low or high and thats why it was'nt working.

Thanks for that as i just now went back and tried it again and i can get a reading now. That will work good if i have to find another one.

I guess i'm gona go try and put this one back on. Hit it with some flux line it up and hit it with an iron. Maybe use some more solder but i'm thinking to play it safe and use whats there and see if i can get away with it as i dont want to push my luck and burn up the resistor or trash the pads.

[ViperJohn]

Thanks for the advice.

Yeah i thought i could'nt get a reading on them untill i took the resistor off.. Prob was i had the DMM set to low or high and thats why it was'nt working.

Thanks for that as i just now went back and tried it again and i can get a reading now. That will work good if i have to find another one.

I guess i'm gona go try and put this one back on. Hit it with some flux line it up and hit it with an iron. Maybe use some more solder but i'm thinking to play it safe and use whats there and see if i can get away with it as i dont want to push my luck and burn up the resistor or trash the pads.

It is pointless to try and reuse that SMR. If you torn the end off is definitely destroyed. If the
solder pad and connecting traces are still intact the SMR can be replaced but I do not know the
value. Never had a reason to desolder it and check. You CAN NOT determine the value to use
by metering it on a know good card while the SMR is still in circuit.

Viper

[KnownKiller]

It is pointless to try and reuse that SMR. If you torn the end off is definitely destroyed. If the
solder pad and connecting traces are still intact the SMR can be replaced but I do not know the
value. Never had a reason to desolder it and check. You CAN NOT determine the value to use
by metering it on a know good card while the SMR is still in circuit.

Viper

A guy on xtremesystems.org told me the value is 35ohms. I've taken a reading on the chip that i have and it says 33-34 ohms.

Welp the prob gets worse then that..

I tried tackin on the resistor on the end with no wire attached and even with flux i could not get it to stick.. Why? Well cause the resistor was chipped on that end so the solder pad had nothing to stick to.. Just bare resistor not the tinned end.

Also.. The worse part that makes this unfixable by me is that fact that now that i've used flux it cleaned the other solder pad spot and i can verify now that theres no pad left on the board..

So now what?

[Spawn-Inc]

It is pointless to try and reuse that SMR. If you torn the end off is definitely destroyed. If the
solder pad and connecting traces are still intact the SMR can be replaced but I do not know the
value. Never had a reason to desolder it and check. You CAN NOT determine the value to use
by metering it on a know good card while the SMR is still in circuit.

Viper

even if there is no power to the card?

A guy on xtremesystems.org told me the value is 35ohms. I've taken a reading on the chip that i have and it says 33-34 ohms.

Welp the prob gets worse then that..

I tried tackin on the resistor on the end with no wire attached and even with flux i could not get it to stick.. Why? Well cause the resistor was chipped on that end so the solder pad had nothing to stick to.. Just bare resistor not the tinned end.

Also.. The worse part that makes this unfixable by me is that fact that now that i've used flux it cleaned the other solder pad spot and i can verify now that theres no pad left on the board..

So now what?

i will go find a old card and try it myself and see if i can't show/tell you a way to help.

btw what flux did you use liquid or paste?

[KnownKiller]

I'm gona post a pic real quick... Looks like a may have a small chance of pulling this off. Looks like i can see where the connection was to that pad. If i'm right can i figure out a way to attach the end of the resistor (new one) to that area where the pad connects to?

[KnownKiller]

even if there is no power to the card?



i will go find a old card and try it myself and see if i can't show/tell you a way to help.

btw what flux did you use liquid or paste?

Paste.. Thats all they had at Radio shack.

[ViperJohn]

A guy on xtremesystems.org told me the value is 35ohms. I've taken a reading on the chip that i have and it says 33-34 ohms.

Welp the prob gets worse then that..

I tried tackin on the resistor on the end with no wire attached and even with flux i could not get it to stick.. Why? Well cause the resistor was chipped on that end so the solder pad had nothing to stick to.. Just bare resistor not the tinned end.

Also.. The worse part that makes this unfixable by me is that fact that now that i've used flux it cleaned the other solder pad spot and i can verify now that theres no pad left on the board..

So now what?

Was that 35ohms with the SMR in circuit or out of circuit. If it is an in circuit reading the number is useless for determining the true value needed for a replacement part.

I can probably repair the card even with the PCB trace damage but I have to know the true, out of circuit, value of that SMR to replace it.

Viper

[ViperJohn]

even if there is no power to the card?

Even with no power on the card. You can not meter true resistor values while they are in circuit
as all other components connected to it will effect the reading you get. What you read in circuit
could be off from actual true resistor value by a factor of 100 or more.

In some rare cases you can get a true value metering a resistor in circuit but you do not know that
if the reading is correct until you meter it out of circuit first. Catch 22 lol.

Viper

[KnownKiller]

Heres a pic.. Best i could get with my camera. Only took 30 pics to get this good of a shot hehe. Its pretty huge weighing in at over 5MBs.

http://home.comcast.net/~eclypse/DSC07157a.jpg

Anyways you can see it better with 10x loop but if you look carefully at the right side of where the pad was you can see where it looks like a trace pulled up with the pad. How can i tell if thats where the connection was to the pad?

[Spawn-Inc]

is that red wire the one you were trying to attach to the resister? if so thats way to big, simple Ethernet wire can be used.
anyway to me it looks fixable, i assume its the one with the tiny bit of solder crossing the white line.



i just tried removing and resolder a small resistor and it was a bitch, i did get it but there is way to much solder which may or may not effect the resistance.


Even with no power on the card. You can not meter true resistor values while they are in circuit
as all other components connected to it will effect the reading you get. What you read in circuit
could be off from actual true resistor value by a factor of 100 or more.

In some rare cases you can get a true value metering a resistor in circuit but you do not know that
if the reading is correct until you meter it out of circuit first. Catch 22 lol.

Viper

i'm not trying to argue, i just want to find out why you can't.

if your MM leads are only touching the ends of the resistor, then shouldn't the current only go to the resistor your touching? though just thinking about that i remembered that there is multiple layers in pcb's and there might be another resistor connected, any truth to that?


EDIT:
i see your point, and just saw the proof.

i tested a resistor and it said 1.0 on the 20k setting, i assume 1000 ohms. then i desoldered it and had to change the setting on the MM to 2000k and it then read 550-580 by itself. after reconnecting it it then read .97 at 20k setting on the MM.

[KnownKiller]

Sapwn-Inc, i did use small wire.. It just looks huge in that pic as its a close up. Its about the size you would get off an old floppy ribbon cable.

[KnownKiller]

Well i have another 8800GTX in the system as i have SLI.. How do i find out the value of the resistor? Do i have to rip the one off the good board? hehe kidding :(

So the 33-34 Ohm reading i got off the resistor while it was not connected is not a good enough reading?

[KnownKiller]

I was checking over the volt mod for a 8800GTS card and the GPu volt mod seems to be identical.. http://www.vr-zone.com/?i=4641#Scene_1


Would'nt it be possable that the same resistor is used for both cards???

I ask cause i bought a fixer uper Evga 8800GTS card thats totally non-repairable and i get a reading of 32 Ohms on the same resistor. Thats a reading with it still connected.

Whats the best way of removing one of these tiny resistors without damaging it.. I'll only have 1 shot at getting this right.

Desoldering Braid or just heat both ends and lift?

[ViperJohn]

I was checking over the volt mod for a 8800GTS card and the GPu volt mod seems to be identical.. http://www.vr-zone.com/?i=4641#Scene_1


Would'nt it be possable that the same resistor is used for both cards???

I ask cause i bought a fixer uper Evga 8800GTS card thats totally non-repairable and i get a reading of 32 Ohms on the same resistor. Thats a reading with it still connected.

Whats the best way of removing one of these tiny resistors without damaging it.. I'll only have 1 shot at getting this right.

Desoldering Braid or just heat both ends and lift?

Possible yes, probably no but it should be close enough. The stock, unmodded Vcore may not be
where it should though.

Their is only one way to desolder SMR's of that size without killing it and that is thermal desoldering
tweasers. The odds are still about 50/50 though.

Viper

[KnownKiller]

Damn man..

Another guy at xtremesystem had an idea HERE (http://www.xtremesystems.org/forums/showthread.php?t=123301) Post 18. What do you think?

[Spawn-Inc]

i would find a piece of steal or copper cylinder and cut the tip like a fork, then bend it so you can hit both sides of the resistor at the same time. this way it should be safe to pull out.

[KnownKiller]

i would find a black steal or copper cylinder and cut the tip like a fork, then bend it so you can hit both sides of the resistor at the same time. this way it should be safe to pull out.

Yeah thats an idea.. If i do choose that route i have a flat tip that i could file a groove into the middle to make it like a fork.

[KnownKiller]

UPDATE:

This morning i had a nice conversation with a guy named largon from xtremesystems.org.

He told me that if i can get the resistor from the 8800GTS card i should solder it where the missing resistor is now by soldering just the one end to the pad (Ground Pad) thats left as the other pad is destroyed.

http://home.comcast.net/~eclypse/solder_points.jpg


Take a small wire and solder that across the 2 RED DOT marked chips and the other end of the replacement resistor. Do this as all these chips are connected and since the pad is gone for the one resistor, this will bridge them.

Then solder the wire from the 1K VR to one of the RED Dot spots to complete the GPU volt mod.

Heres the pic that he showed me with the colored dots and i added the drawing of the wire connecting the 3 resistors.

He also said i could use a 1K VR set to 35 Ohms in place of the missing resistor but i'm not sure how to attach that if i must.

Now this is all going off the idea that the resistor that i tore off is 35 ohms. And that the same resistor used in the 8800GTS is the same one.

Viperjohn, what do you think?


LINK (http://www.xtremesystems.org/forums/showthread.php?t=166324) to the original thread.

[KnownKiller]

I've taken the resistor off the 8800GTS and its all intact. I get a reading of 33.5 Ohms.

After removing the original resistor off the wire to the VR, i get a reading of 34 Ohms solid. looks to be a match.

Now if this mod looks like it makes sense i'll complete the fix. Though i'm not gona have any time to do it till maybe Friday..if not till Monday.

[KnownKiller]

Welp i went over the volt mod area one last time with a DMM to be sure before i start workin on it and i've noticed something.

Check out the pic.


http://home.comcast.net/~eclypse/DSC07157a(1).jpg

The areas with the yellow dots have continuity! So looks to me like that chip on the left is connected to the first ic leg and not leg 6 like the broken resistor and the part to the right of it.

I've also scratched the coating off the area where the broken trace leads to and checked for continuity to the part to the right and yes it looks like the broken resistor and the resistor next to it to the right are connected.

So i was supposed to take a wire and solder it across the 3 resistors.. Well now it looks like i should only do it to the broken resistor and the part to the right only. Which doing so i feel i should go for attaching the VR to the 6th IC leg (Which looks damn near impossable!). I just think thats now my only option as i dont want to mess up the 2 resistors that will be bridged or the soldered wire if i try to touch it again with the VR wire.

[KnownKiller]

I believe i fixed the card! hehe


http://home.comcast.net/~eclypse/DSC07167a.jpg

Well i sat there for a long time figureing out how i wanted to do it. Finally decided to solder the one side of the replacement 33.5 Ohm resistor (From a 8800GTS 320MB) to the one pad that was left. Once i did that i twisted a fine wire together.. I only twisted about 3 strands from the wire so to make it very thin. I then tinned the wire. I then trimmed the wire the length i needed. Long enough to reach from the resistor to the pad where the broken trace lead to. I also had to scratch off the coating on the pad as it was painted. Applied flux to the new pad and held some solder on the pad while i hit it with the iron to tin it.

Once i got the wire to the right length (Took awhile! heh) i applied a tiny bit of flux to both the new pad and the end of the resistor. I placed the wire perfectly so it touched both the pad and the resistor and held it down with a tooth pick. I took my new cheap $7.99 25 WATT "Fine Tip" iron from Radio Shack and held it on the one end of the wire bridge where the new pad was and bam! That baby soldered that bridge right up!

Tested the new connection out with the DMM and all matches exactly what the readings are on the other good 8800GTX.


Oh yeah.. I did the mod this way as i didnt want to attempt to solder the VR wire to that damn small IC leg. Now i can attach it to the resistor to the left.

So in the end all i've done was take the same resistor off a dead 8800GTS card and solder it back on. Only one side was soldered as there was only 1 pad left. Then take a wire and reconnect the broken trace by connecting the wire to the pad that the trace led to and the resistor.

Now back to finally finishing the mod and testing it out.

[KnownKiller]

ITS FIXED!!! ITS WORKING! ITS ALIVE!!!!!! ITS RUNING 3DMARK 06!!! LOL

EDIT: Voltages shown below are taken while the system was in the bios menu.

Memory voltage

http://home.comcast.net/~eclypse/DSC07172.JPG


GPU Voltage

http://home.comcast.net/~eclypse/DSC07173.JPG


Thats default untouched voltage after the mods. I have no idea what the default voltage is supposed to be nor any ideas as to what the voltage i should be shootin for on each, yet.. hehe Time to read i guess.. but first its time to volt mod the other card! :eek: :shakes:

Thanks to all that helped me figure this out!

[KnownKiller]

I guess a pick to clearly show where i soldered the GPU VR to would help those that need to know. I'd say anyone doing this mod should use this spot insted of the other as its a little stronger.

http://home.comcast.net/~eclypse/AltVModSpot.jpg

You could use a 400 Ohm to 1K Ohm VR.

[KnownKiller]

Welp i've been having some crazy strange probs that i've never seen before.

Everything was fine the first day.. so i thought. Started by bumping the voltage to 1.55v on the core and left the mem alone for now but it was sittin at 2.0v.

Found the max GPU Mhz and the Shadder seperate and then played with blending them together. But all hell broke loose on day 2 when i attempted to overclock the memory.

The memory would not overclock at all with everything else overclocked. I tried upping the voltage on the memory to 2.05 just to see if it started coming around and nope! In order to get the memory to overclock from 900 to 950 i had to drop all overclocking and run the card at stock. I was able to hit 1150= 2300Mhz with overclocking the memory alone.

Strange i thought! So i started thinking that maybe i was giving the card more voltage then it wanted to handle so i tried lowering the GPU voltage and see if it made any differenece.. At first it looked like it would still run the high overclock with lower voltage around 1.5v I did have it before to 1.6v.

Ran fine untill i rebooted.. So i lowered the clocks to play witht he ram again. i upped the voltage on the memory to 2.1v and thats when all hell broke loose. Damn 3dmark 06 kept locking up almost right away..

I'm dual booting WinXP Pro and Vista 64 ultimate.. I was just benching in vista first and i almost ended up trashing vista as it kept locking up and i had to reboot so many times. At first it would say there was a prob with the 2nd card in device manager and i could'nt get it working untill i turned the power off totally and powered back up. Also everytime the graphics drivers were gone and i had to keep installing them. I'm like what the hell is going on???

I was thinking something is wrong with the 2nd card untill i would boot into winxp and see that everything was fine. The lock ups kept coming.. Then it was the first card that was in trouble on boot up.

Finally i sat there and thought about what i changed that made this all start.. It was the memory voltage. Wierd cause i thought 2.1v was fine. So i lower it back down to 2.0v and everything started to seem fine again.

I was using winxp from then on cause i thought forsure vista had it and i would have to reinstall that again.. though i wanted to wait til i got the cards figured out so i would'nt trash vista again.

In winxp i finally got my overclocks back on the gpu + shadder. The card was'nt locking up anymore which was good but i still didnt understand why i could'nt overclock the mem at all with the rest of the card.

So finally i lower the voltage down as far as i could on the memory.. all the way down to 1.96v= defalut voltage. Guess what??? I FINALLY COULD OVERCLOCK THE DAMN RAM WITH THE REST OF THE CARD!! lol

Does it make any sense??? NOPE!

Overclocks before with memory at 2.0-2.1v:

core 702-720/ shadders 1782-1834 /mem 900 :(

overclocks with memory at 1.96v:

core 702-720/ shadders 1782-1834 /mem 1150-1200= 2300-2400Mhz


Someone please explain it to me??? Why are my memory not liking voltage?? Are they overheating? Is there a program that reads the memory temps?


Also i've read that people report that the voltage on the memory jumps up like .6 durring load.. I've noticed no movment in the voltage at all. its at whatever i set it to and its solid in idle or load.

Did i do something wrong with the volt mod to the memory??

[nd4spdbh2]

why would it matter if the resistor is in or out of the circuit, i can measure wiht my multi meters all day long the value of resistors and they will be the same in or out of said circuit.

[ViperJohn]

why would it matter if the resistor is in or out of the circuit, i can measure wiht my multi meters all day long the value of resistors and they will be the same in or out of said circuit.


That is so incorrect 99% of the time.

Viper

[KnownKiller]

Welp whatever.. I'm getting tired of messing with it. I can get the memory to run threw 3dmark 06 once overclocked to 1000-1050 along with the GPU overclock but if i go for it twice like back to back it locks up everytime in the first test.

Seems like heat on the memory chips.. cause if i wait a 10-20 mins before letting it go threw again it will be fine.

So if i want a stable system i have to just overclock the GPU/Shadder.

Not a bad overclock though as its able to run over and over and over at 702/1836/900 at 1.58v. I managed a 3dmark 06 score of around 18,400+ in Vista 64. I'm runing a QX6700 at 3.7Ghz FSB at 1648.

I guess it will run me around $200+ if i want to cool the ram better then i am now.

[ViperJohn]

Welp whatever.. I'm getting tired of messing with it. I can get the memory to run threw 3dmark 06 once overclocked to 1000-1050 along with the GPU overclock but if i go for it twice like back to back it locks up everytime in the first test.

Seems like heat on the memory chips.. cause if i wait a 10-20 mins before letting it go threw again it will be fine.

So if i want a stable system i have to just overclock the GPU/Shadder.

Not a bad overclock though as its able to run over and over and over at 702/1836/900 at
1.58v. I managed a 3dmark 06 score of around 18,400+ in Vista 64. I'm runing a QX6700 at 3.7Ghz FSB at 1648.

I guess it will run me around $200+ if i want to cool the ram better then i am now.

Drop the Shader clock 54Mhz and lower the Vcore down to around 1.500 or less and you will
probably see the memory clock again.

Viper

[KnownKiller]

Drop the Shader clock 54Mhz and lower the Vcore down to around 1.500 or less and you will
probably see the memory clock again.

Viper

Yeah thats about where i was when i was trying to get the ram to clock.. i had the gpu at 684/1782 though i kept the voltage on the gpu the same at 1.58v. Since the ram aint stable i just tossed it out and clocked the core back up.

I may try again tomorrow with lowering the gpu voltage to 1.5 and see what happends.

Its just a dumb case design on my part. I have the 2 120.3 PA rads blowing all that hot air right onto the cards.

Only thing i could do to eliminate that prob is to mount the rads on the back and top of the case.

I tell ya i'm liking that new Swiftech stealth block more and more everytime i look at it. Calling my name hehe.. Dont want to pull the trigger though cause i'm sure to score new cards soon after the new year.

[El<(')>Maxi]

Good job known, thats not an easy fix and you nailed it.

[KnownKiller]

Thx man! Yeah i just should be happy its all working then caring about the overclocks hehe.

[xilix]

Thx man! Yeah i just should be happy its all working then caring about the overclocks hehe.

Exactly. Really puts it in perspective when you lose a part that would be expensive to replace. I stopped OC'ing my GTS where you see it now because I just can't afford to replace it. I'm happy enough I have an 8800, if I lost it I'd be devastated, so I take it easy.

CONGRATS on the fix, man. Seriously. I'm happy to see you got it back up and running.

[KnownKiller]

Well i've finally got some time to upload some pics for ya guys. Before overclocks i was getting in around 17,500 in 3DMark 06.. I was hoping to break 18,000 heh.. Blew that away.

Heres what i'm keepin it for now at 1.6v/2.2v <-- I will lower the voltage on the ram today as i'm sure i dont need that much.. though i remember runing the ram at 1000Mhz and 2.10 didnt seem like enough.. though it could of been something else wrong like rivertuner and the drivers not playing right.


http://home.comcast.net/~eclypse/3700MHZ-684-1836-1000-winxp-YES!.jpg


Heres the highest stable 3dmark i could get with no artifacts..

http://home.comcast.net/~eclypse/3700MHZ-684-1836-1050-winxp-WOW!.jpg


Heres a shot at the highest core i could get stable at 1.6v along with the shadder and mem clock.. Basicly the highest each i could get to blend well.

http://home.comcast.net/~eclypse/3700MHZ-702-1836-945-WinXP-NICE!-3.jpg


I've had a hell of a time figuring out the proper way to overclock these cards with so limited info i could find online. I did have the card runing over 720-729 core but that was early on as i though i had to set the voltage under load! lol.. I set the core voltage at 1.6v under load so it really could of been 1.67+ Temps were 45c idle and 51c load. Figured it was like when i did the colt mod on my former 7800GTX cards as the coltage would'nt kick in till it was runing 3D.

So i doubt the ram is overheating much as its runing fine right now at 1000Mhz with 2.2v on them. I still would love to go with the Swiftech Stealth.. maybe hit the parents up for a X-Mas gift!

So early on i had probs with i believe rivertuner not overclocking both cards as i never saved the timings and rebooted.. That plus i've heard some say that the newest 192+ drivers dont allow overclocking or dont atleast on the ram. So everything was peachy once i switched to the newest offical drivers and stopped using rivertuner and just made a crap load of custom bios for the cards with different timings.

It was a sweet setup bios flashin with 684/1620/1000 - 684/1736/1000 - 684/1836/1000 ect.. ect.. All with different ram Mhz 900/945/1000/1027/1045/1080 hehe. No them all by heart after all that work.

I think 684 must be the sweet spot though.. I could'nt go higher then 702 with 1.6v stable though.. Would love to try 729 as i've heard someone say that the card really kicks up then.

Oh and vista's benches are about 1000 points lower.

I've played COD 4,UT3 and Crysis and all seems fine! Though i cant get crysis to run in 64bit mode in Vista 64 so i'm gona try reinstalling Vista today since all those lockups from testing the overcocks must of messed it up.

Also i would love to hit 19,500+ or 20,000.. Any tips? Never tweaked 3dmark before and i'm have not tweaked any settings for the drivers or vista or winxp.

[KnownKiller]

http://home.comcast.net/~eclypse/DSC07209a.jpg


Quick pic of the 8800GTX SLI in the water loop.. Maybe this weekend i can finally show off the new custom mountain mods UFO case.. hehe. If i manage to i'll post a thread on it and also link it up in my signature.

Ahhh what the hell i'll post a quick tease since it shows the 2 rads and pump mounted in this shot.

http://home.comcast.net/~eclypse/DSC07036a.jpg