Running 2x R9 280x causes BSOD with error code 3b

I have this issue whenever I try to install two R9 280x's at the same time - I got three and they're all from different companies. Whichever constellation I install them in, once I insert the second card, it does the BSOD with the 3b error code (driver related) or a complete crash with no error message or system dump.
Apart from this, the three different R9 280x's work, except the MSI card won't show properly in CCC, still shows fine in the device manager. The Gigabyte and the ASUS work fine when solo . . .
So basically I need to figure out a way to run 2+ of these cards at the same time, without the terrible BSOD or any other kind of crashing...

My system:
CPU: i3-4130 - no OC.
MB: Asrock Z87 Pro4 - flashed to newest bios 1.70.
RAM: 2x16GB Sniper at 1600Mhz (auto setting, but also tried at 1333Mhz).
OS: Windows 7 x64 Professional. Fresh and fully updated.
APPZ: Avast Antivirus, CGMiner 3.7.2, various monitoring tools and other small professional programs - I don't run these when testing though!
FURTHERMORE: One 500Gb Western blue and one ASUS DVD reader/burner.
GPU's and DRIVERS: ATI 13.11 Beta 9.5 drivers and tried others.
Onboard HD 4400 which I use for the desktop.
MSI R9 280x Gaming - got the new vBios 015.039.XXXXXXXXXXXX
ASUS DirectCU II TOP R9 280x - 015.039.XXXXXXXXXXX
GIGABYTE R9 280x Windforce OC version - think it's 015.38.XXXXXXXXXX
PSU: Toughpower XT Gold 1475w

I get the crash before I get a chance to log into Windows or immediately after.
I tested it with any two cards yes and with cards in both the x16 and the x4 slot.
The computer starts up fine in safe mode, where I find the two cards in the device manager (tried Gigabyte and ASUS and MSI and ASUS) and I haven't been able to replicate the BSOD in safe mode.

I tried the following:
1. re-installing cards in various constellations and slots, with driver-sweeper removing drivers completely and then re-installing. I've confirmed that both my full sized PCIe slots work with cards individually.
2. Flashing bios to newest version.
3. memtest86 and MemtestHCI, then I disabled the page file and ran both tests again - one full pass for each and running four isntances of MemtestHCI - I also tried with 2x2Gb GSkill, before upgrading to 16Gb RAM.
4. I tried 13.11 Beta 8, Beta 9.2 and Beta 9.5, with the same results and using driver sweeper in between to fully remove the earlier version.
5. Running in safe mode (with 2 cards), after using CCleaner to make sure almost nothing runs at startup... Worked fine.

I was thinking about doing the following:
1. Upping vcore, even though I'm at the moment running everything at stock speeds (no overclock) and I fail to see why it would help.
2. Flashing the video cards bios, but not sure if I can find never versions or if I can use e.g. an ASUS vBios in a Gigabyte card and the other way around.
3. Trying out other older drivers - there aren't many left that I haven't tried though.
4. Even though this is a fresh install, I was thinking of making another fresh install of Windows 7 x64, with all updates etc and the necessary drivers to run everything, except the GPU (which I will wait with), then make an image of the file, so I can get back to a clean machine quickly, before I install GPU drivers. Finally install the GPU's - thus doing the drivers and basic installation before anything else - and I mean absolutely anything else . . .

I sincerely hope someone can help me, as I am at a complete loss here . . .

Hope anyone can help...

EDIT: updated what has been tested and done.

 

ASUS TOP strikes again. lol

3B is officially a video driver problem, in my experience it's usually triggered by the card itself.

Reduce the core/vram clocks and see how that goes......otherwise try drivers.

Do u have a dumplog?

 

Thanx for the reply...
However, I'm running the cards at stock speed or rather the one that has been installed will have been running fine solo, then when I add a second card, upon starting up, Windows throws the BSOD and error before I get fully into the OS and thus it isn't a problem I get from doing an OC on the CPU or GPU. I ahve tried to OC the clock speeds a bit on each card, but that didn't result in this BSOD 3b error - only starting up with two cards did this . . .
I could perhaps lower the core and memory clock of the card already in the computer, but I fail to see how that would help.
I also thought of lowering the CPU voltage, but it's at default - should be stable...
I can produce a dump log...

 

Dump would be handy.

 

I tried all three cards individually again and noticed that the Gigabyte also has the issue with not showing in anything but GPU-Z and with the temperature in Open Hardware Monitor (also in the Device Manager). Not in CCC either...

Anyways, got a dump, but I can't post it, as I haven't got more than 10 posts - in fact this is only my 5th!

 

Sys dump:
http://www.mediafire.com/download/m9euu0u52m621ug/121113-24866-01.dmp

 

Got any others? or Event Viewer?

Make sure to run the cards separately....and u should bench or stress the abnormally behaving GPU (Gigabyte).

Can u please reset everything to default, no overclocks at all.

 

nvm..

 

Another:
http://www.mediafire.com/download/bg3qvd3e3zv6wua/121113-46472-01.dmp

 

I've got only 2x2Gb of RAM, though I'm running Win 7 x64 pro.
I was thinking maybe I'd need more RAM and I could put 16Gb RAM in it just to test, it's just hard to see why this would be an issue - I've run 3 and 4 GPU's in other setups and with as little as 2Gb RAM.
To answer your quesion I have 3760Mb of virtual memory and I figure that's in a pagefile somewhere . . .

EDIT: Just read that it's minimum 4Gb for one of them and 8Gb recommended for crossfire, so will try that no - 16Gb should rule out any issue of it being not enough RAM...

 

^ok.

I have yet to analyze the dumps but I will,

More questions;


anything in Event Log?
What were you doing when the crash happened
Is memory remapping enabled in BIOS?
Does it happen with any 2 cards?
does increasing tghe PF size help?


Btw a guy posted here a fgew weeks a ago with same problem u have.

 

I know that post I think, but his problem or rather his BSOD happened when gaming and after successfully starting up - I get the crash before I get a chance to log into Windows.
I tested it with any two cards yes and with cards in both the x16 and the x4 slot.
I installed 16Gb of new RAM, which increased the PF size to 16048Mb, but the BSOD with same error persists.
The only issues in the event viewer seems to be this crash and recovering from it . . .
New sys dump.
http://www.mediafire.com/download/pem1d4w91b9bv87/121213-23056-01.dmp

I appreciate the help very much!
Where I'm at there aren't really any shops or the like that can offer any help . . .

 

Oh wow, lol. I misread your first post and took it as "I have tried older drivers" instead of "thinking of trying".

Yep definitely try different drivers mate.....

your dumps say it's a driver error, but also reference a memory access violation which is why I said memory. Access violations can be caused by corrupt drivers though.

seems a bit strange to be drivers since no one else has reported the same issue, but hey u never know.

I also see Avast is running, can u disable it? In fact disable any monitoring tools as well.

Really u shouldn't have more than one open at the same time anyway. When multiple apps probe the same sensors instability can result.

 

Ah wait.....are you also running a bitcoin miner? If you are, please disable this first before anything else and test again.

If it still fails, then u can try a different set of drivers.

 

I'm not running CGMiner for the purposes of this at all!
I used it while running one card only . . .
I will be running it once the cards stop acting up though.

I tend to use only Open Hardware Monitor, but when that fails I try out others and as such I also installed GPU-Z, HWMonitor, CPU-Z, CCC at times and MSI Afterburner, but none are running at start-up and now I've also disabled AVAST for the time being.


With the ASUS card installed first, I then tried the Beta 9.2 drivers and they let me log into Windows, after installing a second card, but once I tried to check the card (device manager one time and a monitor program the second) it immediately crashed - BSOD with 7e and this is the system dump:
http://www.mediafire.com/download/ma2mj2umbevkb08/121213-50388-01.dmp

I also tried with the Gigabyte, which resulted in a crash, but without the BSOD - just black screen, then boot up again and repeat. I can't find anything in the event viewer about this...

I will try the MSI again and see if it reproduces the error and then with the Gigabyte again - this time without AVAST or anything else running for that matter...
If they keep crashing, I'll try the Beta 8 driver instead...

 

Do u know how to uninstall any hidden devices in Device Manager? Can u do this and try running driver sweeper 3.1.

Just want to make certain it's not drivers because everything I'm looking at says drivers.

I might have to do an analysis your dump...which means I need the symbols..grrr.

 

Update

Do you have a RealTek LAN controller on your board? What happens if you disable it?

 

I installed beta 8 and reproduced the errors - BSOD with 3b error.

First card in both tries was the ASUS.

When inserting MSI I got this dump:
http://www.mediafire.com/download/h6rl85b6cs8wa0e/121213-22339-01.dmp

The Gigabyte produced this one:
http://www.mediafire.com/download/gyxxgux7gk8kanh/121213-21200-01.dmp

I got an Intel Ethernet Connection I217-V for my ethernet, though it's not plugged in. I also have a Tenda USB wireless LAN card, but I can unplug it, then run driversweeper, install new drivers and check. However, I also ran driversweeper prior to the test I just did, with Beta 8 . . .
I have a Realtek High Definition Audio driver, but nothing else from Realtek.

EDIT: I've been using DDU - Display Driver Uninstaller - I'll find the one you're using and try uninstalling Beta 8 completely, then re-installing with Beta 9.5 divers and test with two cards - all of this with the USB wifi unplugged . . .
Not sure if I should do the new drivers or older ones, but figure try new ones first...

EDIT: shouldn't I just take 3.2 Driver Sweeper?
Well, will try 3.1 first anyhow, simply because you said so

 

^Oh hell my bad, I was thinking 3.1 was the latest... 3.2 is fine. lol


Anyway, while having a closer look at your dump it dawned on me that I've seen this before.

Run memtest86 AMD memtestHCI for at least 5mins reduce your RAM to 1333mhz and disable the pagefile. (This is to eliminate RAM and poss bad HDD from the picture).
Open memtestHCI 4 times so all RAM is tested, not just some of it.

THEN assuming memtest passes, please uninstall all the Realtek drivers and delete everything with driver sweeper especially rtkvhd64 .

Also suggest running CCleaner but make sure to untick the memdump entry. CCleaner isn't a "registry cleaner" btw.

Retry.

 

I've set my RAM to 1333Mhz and disabled the page file - before I disabled the pagefile, I ran 9 tests with Memtest86 and a full run of 2048Mb paired using MemtestHCI and now that I've disabled the page file, I'll run both memory tests again, but with 4 x MemtestHCI taking care of all of it.

Prior to running memory tests, I uninstalled anything relating to Realtek, which meant only Audio, but still. I used Driver Sweeper and also removed all AMD display drivers. Finally I manually deleted the rtkvhd64.sys file . . .
The BSOD didn't dissapear, so I've done these memory tests and now will do them again, having disabled the page file . . .

If there aren't any memory errors, I'll take a look at CCleaner . . .

I'm super tired, so having a hard time following instructions