• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

what is wrong with amd

directly from the user manual:

slot1: PCIe 3.0 x16
slot2: PCIe 3.0 x8_2A
slot3: PCIe 3.0 x8_2B
slot4: PCIe 3.0 x16/x8_3
slot5: PCIe 3.0 x8_4

suggested setup for sli/cf:

single VGA: slot 1
sli/cf: slot1 - slot 4
3way sli/cf: slot1 - slot2 - slot3
... etc.

so no.

I'll go ahead and run memtest but i doubt that's the case since i had the same issues on my other system which had new ram as well (what are the odds that both kits have a defective peace ofr am in them?)
 
Weird, I have NEVER seen a motherboard like that. Try top 2 and see what happens, can't hurt anything. Every CF/SLI setup i've used has required consecutive slots............not slot 1 / 3, always top to bottom.

In fact, the P8Z77WS which is also a quad SLI motherboard states explicitly to use the top 2 slots for crossfire / SLI on 2 cards, even though the 4th slot down is a single pciex16.

Try the top 2 out, that may help.......
 
Last edited:
xoleras said:
Weird, I have NEVER seen a motherboard like that. Try top 2 and see what happens, can't hurt anything. Every CF/SLI setup i've used has required consecutive slots............not slot 1 / 3, always top to bottom.

In fact, the P8Z77WS which is also a quad SLI motherboard states explicitly to use the top 2 slots for crossfire / SLI on 2 cards, even though the 4th slot down is a single pciex16.

Try the top 2 out, that may help.......
Click to expand...
Yeah, this certainly wouldn't hurt.

And killernerd, I had a defective stick of RAM straight out of the box from Corsair once. They replaced it no-hassles under warranty so I'm not disparaging them, these things happen (and I've bought 3 Corsair kits since with no defects). But bad memory can cause very strange things, BSODs, etc. Especially if it's only a single stick out of four.
 
Your cards are fine in the slots they are in if that is what the manufacturer says. As far as the power supply, I'm still gonna suggest you try it. If you actually have a 850 watt psu lying around and are too lazy to try it then honestly bro, I'm sorry to say that you deserve what's going on. /shrug

Keep in mind you don't even need to put the psu in your case. Simply open your side panel and run the cables to where they go. As far as your ram, I'd agree that since you had the same thing going on before on the other setup and this is a new kit of ram that it's very unlikely to be the ram.
 
xoleras said:
Pretty sure you have to use the top 2 ....

Thats the way every motherboard i've owned is, the slots closest to the CPU have the most pci-e traffic....
Click to expand...

No. He is doing it correctly. The second slot is an x8 slot. The third red slot is the other x16 slot.

xoleras said:
Weird, I have NEVER seen a motherboard like that. Try top 2 and see what happens, can't hurt anything. Every CF/SLI setup i've used has required consecutive slots............not slot 1 / 3, always top to bottom.

In fact, the P8Z77WS which is also a quad SLI motherboard states explicitly to use the top 2 slots for crossfire / SLI on 2 cards, even though the 4th slot down is a single pciex16.

Try the top 2 out, that may help.......
Click to expand...

SLI has never required consecutive slots. Some SLI bridges mandated where you could put the cards, but that was by limitations of the SLI bridge length, nothing else. The motherboard's "correct" slot configuration for SLI and Crossfire varies, but is usually in line with the standard SLI bridge. And for most of them, they alternate. I've always seen them be 1 & 3 and 1, 2, and 3 for 3-Way sLI.

Lord_Exodia said:
Your cards are fine in the slots they are in if that is what the manufacturer says. As far as the power supply, I'm still gonna suggest you try it. If you actually have a 850 watt psu lying around and are too lazy to try it then honestly bro, I'm sorry to say that you deserve what's going on. /shrug

Keep in mind you don't even need to put the psu in your case. Simply open your side panel and run the cables to where they go. As far as your ram, I'd agree that since you had the same thing going on before on the other setup and this is a new kit of ram that it's very unlikely to be the ram.
Click to expand...

Archaea said:
EXACT same thought I had after the first few lines of this saga.
Click to expand...

Again, it's NOT the PSU. An insufficient of failing PSU does not manifest with this type of problem.
 
Rizen said:
Yeah, this certainly wouldn't hurt.

And killernerd, I had a defective stick of RAM straight out of the box from Corsair once. They replaced it no-hassles under warranty so I'm not disparaging them, these things happen (and I've bought 3 Corsair kits since with no defects). But bad memory can cause very strange things, BSODs, etc. Especially if it's only a single stick out of four.
Click to expand...

I am aware that it could be a defective stick. I am, however, doubting the possibility that 2 different kits both have a defective peace.
I bought a new ram kit together with the new psu and the 7970's. Now that i build a new system i, once more, bought a new ram kit (4x4gig).

Both systems give roughly the same crashes that's why i'm doubting the defective ram theory.
i'm going to run memtest nonetheless just so i can say that i really did try everything.
 
Yeah, I know, it's unlikely. But I thought the same thing too with my system a few years back, yet lo and behold.. I do acknowledge since you had TWO different sets of memory with the same problem that it's likely not the RAM, but running memtest is easy and you won't have to tear down your system to do it. :)
 
I have run into several cases of xfire 7970's not working using 12.4 drivers. I'm not reading through all the posts to see if you've tried a different driver version.
 
Don't know if this has been posted, but when "hardware acceleration" on flash is enabled, it messes with 2d/3d clocks and voltage. Right click on the youtube video and disable hardware acceleration. You may also try matching 2d clocks to 3d clocks with a tweaking utility of your choice and see if that helps.
 
killernerd said:
I am aware that it could be a defective stick. I am, however, doubting the possibility that 2 different kits both have a defective peace.
I bought a new ram kit together with the new psu and the 7970's. Now that i build a new system i, once more, bought a new ram kit (4x4gig).

Both systems give roughly the same crashes that's why i'm doubting the defective ram theory.
i'm going to run memtest nonetheless just so i can say that i really did try everything.
Click to expand...

All of this stuff is minimal effort so there's no harm trying. It takes little time to run 1 RAM stick and changing pci-e slots......who knows. I also have had a bad RAM stick at one time unrelated to my GPU's, but I randomly decided to run 1 stick of RAM after 4 hours of troubleshooting. Its worth a shot.
 
Try changing "PCIE_x16_1 Link Speed" from Gen3 to Gen2 in the BIOS if you haven't already. SB-E chips do not official support it, and some can't handle it well enough and it can result in BSODs every now and then.

Another thing to try is going into your power management profile in windows, and change "PCI Express -> Link State Power Management = OFF." It can't hurt to try it.

If it still has problems I'm thinking it's the drivers, since you're able to run one card at a time fine enough.

You can view the BSOD details with this program, assuming your system is set to save the dump files.

The next thing that could be tested is the RAM. Though it seems that you got a new set of RAM with your upgrade, which does make it seem strange to have the same errors as your old system. Stranger things have happened I guess. You can first try it without the XMP timings and run it at its auto default values and see if that helps to any degree.

Run memtest86 as long as you can and even if it runs without errors, the next step is to run as many instances as needed of HCI Memtest from within windows. Sometimes memtest86 doesn't trigger things that would otherwise happen in real workload environments. Just run the HCI instances until they reach at least 500% completion, with 1,000% preferred.
An other option is Y-Cruncher which will stress anything that isn't a video card.

It's weird, but I've had a RAM kit that passed 30hours of memtest86, but it sure as hell wasn't stable in windows for even 5 minutes.
 
I'm an AMD/ATI fan and I'll tell you this straight up, they aren't going to make it easy for you. If you want good compatibility go with Nvidia. You might have a little more luck not getting BSOD's.

I have an HD6850 right now, but in the past i've had a dozen ati cards and nvidia cards. Here's what you have going against you: New hardware, hardware that a small percentage of people have, and multi-gpu.

If you had an HD5770, or 9800GTX or HD5870 or GTX560, or some other hugely popular card in single-gpu setup, you would not have nearly the same number of problems. As you said, "Fucking shit, man." is right.

Your best bet now is ATIman uninstaller with frequent reboots, CAP profiles, RC11, or test driving various driver + cap combinations to suit a single gaming application. You are in the worst boat possible right now with 7970 Xfire. Even the 680 and 670 have some problems with stuttering and vysnc right now, but I expect Nvidia to iron out their red flag issues much sooner than AMD will.

Good luck
 
nvidia are alot better than amd/ati when it comes to stability, especially driver and "out-of-the-box" stability.

as the poster above said if you want something that works straight-out go with nvidia.
AMd may have cheaper prices and better performance but it does come at a cost.
 
alright, i'm taking in everything that has been said here. Allow me to give a small update.

Memtest came out clear, surprise surprise.
the bsod viewer says that the bsod was caused by dxgkrnl.sys, which apparantly links back to the amd drivers (again, what a surprise).

On bootup the post screen says that it's running the pci-e slots in "gen2 mode" which would be pci-e 2.0 mode i presume? Even though they are set at gen3 mode when i look at the bios settings... Strange. i am aware that the current chips do not support gen3 "out of the box" but have some sort of magic going on that forces them to work with it nonetheless.

Just to be sure i'll set the bios to gen2.

Next up: html 5, i hate it. It's really buggy, at least for me. Videos won't go in high res, they just stop loading and playing. I have to play them in lower res. Refreshing sometimes helps but w/e that's a whole other topic.
I disabled hardware acceleration for flash and set the "local storage size" to unlimited.

doing that seems to have solved the issues i had with it so far. no more "out of memory" exclamation marks and so far no more crashes.

I have been testing some of the head ache inducing games that would crash before and have yet to freeze. (dirt 3 and anno 2070 for instance)
fingers crossed that it keeps up like this.

conclusion: maybe the culprit was the drivers all along? damn you amd T.T

Not that i'm entirely convinced that it was a driver error only. My old pc would crash and freeze far more often. So perhaps it was a bit of both on that system.
After all it did fry my 4890's...
 
speshial said:
nvidia are alot better than amd/ati when it comes to stability, especially driver and "out-of-the-box" stability.

as the poster above said if you want something that works straight-out go with nvidia.
AMd may have cheaper prices and better performance but it does come at a cost.
Click to expand...

If the TDR issues and GTX 680 RSOD was not around then you may have a point.

Owning 1x 1800xt, 2x 1900xt, 4x 3870, 2x 5970 and 3 screens a set of drivers has never given me stability issues as far as BSOD or driver has stopped responding, i have had the occasional Driver/CAP that would make a particular game CTD but that is the exception.

And i think you mean "New out of the gate" NV are better than AMD and not "out-of-the-box" as if every gfx card needs a cure time when opened.
 
Last edited:
speshial said:
nvidia are alot better than amd/ati when it comes to stability, especially driver and "out-of-the-box" stability.

as the poster above said if you want something that works straight-out go with nvidia.
AMd may have cheaper prices and better performance but it does come at a cost.
Click to expand...

LOL do some research, before you make embarrasing statements like this.
 
You must log in or register to reply here.
Back
Top