New SLI/surround setup: Issues and Questions.

El Nacho

Noobie Cheese
Joined
Jan 22, 2002
Messages
2,258
Get the easy questions out of the way.

Which GPU is "gpu1"? I assumed the top one was, but the nvidia control panel calls the top one number 2. Temps and load on the cards while playing games indicate the top one is 1. Issues below assume afterburner is calling the top card gpu1.

Is it normal for both cards not to be at 100% load? played the just cause 2 demo loads range from 100/70 to about 85/85.


Issues.

heat is the big one. I first and currently have the cards setup with 1 in the top slot and one in the lowest slot. I did this to give the cards as much space as possible to keep cool. During games the cards get up to between 85 an 90 c. I found out that the slot the second card is in is only 4x, I moved it to the third slot. This caused the top card to spike to close to 100c in games while the second card stayed around 75. Both times I used msi afterburner to force the fans to 100% after 80c

So should I return the cards for one with better cooling or am I looking at not enough system cooling?

The other issue is that my middle monitor turns off any time windows asks for permission to do anything. I have to turn off surround and back on to get it working again. Ideas?


Specs:
I7 920 @3.4
6gb corsair xms
evga sli le
2x palit 460

Temps idle:
CPU @ 58 c (stock hsf)
system @ 35 c
gpu1 @ 44 c
gpu2 @ 53c
 
Last edited:
GPU1 is decided by the driver based on some arbitrary thing.. it doesn't matter to you though.. they could be in either order.. it's just a name. The driver doesn't necessarily know which slot is the "top" one physically so...

Don't worry about load, it will do what it needs to. There is no way to get totally accurate "load" measurements on a GPU... I don't know why all these tools try to make it as if it's as simple as CPU-load.

The boards are designed to go up to 100C. Of course cooler is better, but don't expect issues until 105C. That's when they'll downclock to avoid frying, and if they progress to 110C anyway somehow, they'll shut off entirely. So there is protection in place. But that said, the GTX 460 is known to be cooler than this.. read some reviews on it and see what people claim it gets up to under load (I'm not sure) for comparison.

Ideally new Fermi boards should be kept at 95C or 90C or something max. 80C is totally overkill.. no need to cool it that far.

The middle display turning off is weird though... do you mean when the UAC popup grays your screen and has the one accept dialog?

What 3 displays are you using? Are they all identical?

BTW you should be using the first and second PCIE 2.0 slots.. the first one is 16x and the second one is 8x. The third one should be 4x.. That's weird that you got 4x in the second one. See your product manual here (i assume you have the x58 LE?): http://www.evga.com/support/manuals/files/141-BL-E757.pdf

Look for the page called "Expansion Slots":

1 – PCI-E 2.0 x16 - Slot 1 (Primary)
2 – PCI-E x1 - Slot 1
3 – PCI-E 2.0 x8 - Slot 2
4 – PCI - Slot 1
5 – PCI-E 2.0 x4 - Slot 3
6 – PCI-E 2.0 x8 - Slot 4
 
should clarify that a little better. the best temps i've gottne are in slot 1 and 5. But that gives me the issue of the second card only in a 4x slot. Putting them in slot 1 and 3 give me the close to 100c on the 1st card.

Most reviews/forum posts ive read all say the cards stay around 75.


UAC is what cauases the monitor to say input not valid and all 3 monitors are brand new, same model.
 
What is causing your GPU to get to 100C? GTX 460?!?

You must be running the things in a cardboard box.. those aren't hot cards.
 
should clarify that a little better. the best temps i've gottne are in slot 1 and 5. But that gives me the issue of the second card only in a 4x slot. Putting them in slot 1 and 3 give me the close to 100c on the 1st card.

Most reviews/forum posts ive read all say the cards stay around 75.


UAC is what cauases the monitor to say input not valid and all 3 monitors are brand new, same model.

Right but I asked what monitors they are. There's no reason that should cause a loss of signal..
 
Right but I asked what monitors they are. There's no reason that should cause a loss of signal..

acer g235h. Could just be a bad monitor or cable. its just really odd that it only happens when uac pops up. I'm going to switch the cables and then the dvi ports to try an narrow down the cause.
 
What is causing your GPU to get to 100C? GTX 460?!?

You must be running the things in a cardboard box.. those aren't hot cards.

this was during the just cause 2 demo and the cards in slot 1 and 3. The case is an older lian li nockoff. Two 80s in the front, two in the back, 1 on top.
 
Are the fans running? Maybe you turned on user controlled setting in MSI Afterburner instead of having them on auto? Or maybe your kid stuffed Play doh into the exhaust ports?
 
Are the fans running? Maybe you turned on user controlled setting in MSI Afterburner instead of having them on auto? Or maybe your kid stuffed Play doh into the exhaust ports?

I used msi afterburner to make sure the fan is going to 100%.
 
Think i'm going to take them back and get the msi version. See if that helps with the heat.
 
That is definitely too hot for 460's. I have over clocked 480's and my max temp's are around 75 C under full load in SLI with them right next to each other. I have my fans scale much earlier though in MSI Afterburner. 100% fan kicks in at like 65 C, super loud but it doesn't bother me as my computer is physically in another room from my desk.
 
they didnt have the msi in stock, (though I may still end up with it) so I got the upgraded palit with the better HSF and grabbed a 690 II advanced case. these cards do run cooler than the others so max was only 90. With a fan in the door blowing over the cards it stays around 72.

I'm going to have to assume any with stacked cards has an external fan helping out.
 
Back
Top