Dual ATI Radeon HD 5970's (Alienware Area-51 ALX)

That's exactly what I'm saying, Chris; P19 and p20 first card; second p18 and p21 --doesn't work. As soon as I get into Crysis, it automatically reboots my computer; along with Starcraft 2 and FFXIV. But when I have other variations of the rails such as: P17 & P19 -- P21 & P24, I'm able to play games, for random intervals before it either freezes and gives me bsod, or just restarts my computer by itself.

I just got my graphics cards, and after hooking them up (tried the chart once more) it restarted my computer upon entering Crysis. But after I switched the rails around, I was able to play... However my graphics cards almost reached 90ºc for the short amount of time I was playing. I havn't crashed yet, but I JUST got them... I just feel like something is wrong with the temperatures considering it's way different from other people's idle and max load...

I've tried all your combinations with support on the phone, aside from the second to last one because I have a creative soundblast card there. And I get the same results...

Radeon HD5970 Vbios to correct several issues. You may see any of the following -
* Blank (black) screen
* Blue Screen of Death (BSOD)
* Artifacts
* Display driver stopped working error message
* Video distortion
* System hanging or lockups
* Screen tearing

Notice - Do not shutdown or restart the computer during the VBIOS flash process. Doing so may permanently damage computer hardware.

* Download the Vbios R277188.exe to your desktop
* Once downloaded, double-click R277188.exe. It will extract to C:\dell\drivers\R277188
* Insert a blank CD-R or CD-RW in the optical drive or a USB flash drive into a USB port. If there are two optical drives in the computer, ensure the optical drive that is to be used has the capability to write to discs. This utility is only compatible with CD-R or CD-RW disc media, not DVD or BD media
* In the Alienware Boot Media Creator window, click to select either Install to a USB Flash Drive or Create a Bootable CD
* Depending on the medium used, use the appropriate step below:
- On the Install to a USB Flash Drive window, select the USB flash drive from the drop-down menu and click OK to create the bootable flash drive
- On the Create Bootable CD Image window, ensure Burn the image to CD-R or CD-RW media instead of saving it to a file is checked, and click OK. On the Burn CD window, select the optical drive from the drop down menu with the disc and click Burn
* Once the file has been completely burned to the disc or written to the flash drive, leave the disc in the optical drive or the USB flash drive in the USB port
* Click Finished Creating Media
* Click Start, then select Restart
* At the Dell or Alienware logo screen, press the F12 key until F12 Boot Options is highlighted in the lower left-hand corner
* On the Please select boot device screen, select the optical drive with the recently burned disc or select the USB option and press the Enter key
* On the CAUTION! screen, press the B key to continue
* Once the VBIOS flash tool is complete, the message You may safely shutdown the computer will appear. Remove the disc from the optical drive or the USB flash drive from the USB port and restart the computer
* Once the computer has completely restarted, right-click the Windows 7 Desktop and click Catalyst Control Center
* In the Catalyst Control Center window, click Graphics and then click Information Center
* Click to select the Graphics Hardware tab
* Verify the BIOS Part Numbers match the tables below:
- Dual Video Card
Primary Adapter 113-C00001-104
Linked Adapter 113-C00001-104
Linked Adapter 113-C00001-105
Linked Adapter 113-C00001-105
- Single Video Card
Primary Adapter 113-C00001-104
Linked Adapter 113-C00001-105

The first Alienware Tech that I spoke with, already did that via alien contact. I have the disk in my leather bound book you guys place with our machine. I checked up on the table you listed and everything matches in Catalyst.

Hello,

I had the very same problem, and morblore, here in that forum, helped me fix this issue.

You can read the following thread :

true

To resume the thread, I could at last use my computer when connecting every connectors on the cards with a different rail.

4 connectors <-> 4 rails.

I really think the issue is the cascading cable : All the power goes through the first connector, then sent to the others, instead of being distributed correctly since the very beginning.

To answer Chris :

Are you saying that the HD5970 on this chart is not correct?

In my case, the chart is obviously wrong, it does not seem normal to have the cables in this way (see the picture in the thread).

By the way, despite I didn't check recently the temperatures, I think I still have the same temperatures as yours...

I hope these information will help you a bit.

Tristan

I'm currently using Primary: P20 & P17 and Secondary: P21 & 19

I was on the phone with another tech rep yesterday, and he agreed that the temperature my GPUs are currently at during idle and load are unusual. Generally people have a 50-60ºc in idle, and 70-80ºc max load not 70-80ºc idle, and 95-105ºc max load...

The tech said he was going to have a machine built with the same exac specs as my current computer and try to replicate the problem and fix it from there --however I don't see how this will work...

Edit: Also, Tristan, I used your thread before all this nonsense to try to see if any of this worked, and it does seem as if I can play my games, but I want to fix the temperatures on my GPU --I don't want to melt my hardware.

Video card primary = P? and P?
Video card secondary = P? and P?

Instead of searching for a solution, the LAST tech that I talked to (Just a few minutes ago) just enabled manual fans and made it noisier... This is sort of irritating for a $7,500 purchase.

Try each card by itself in the primary slot. What are the temps?

60ºc-70ºc idle, and 92ºc-102ºc load

It fluctuates --this is very irksome! Everyone else says it's supposed to be 50-60º idle, 70-85º max load! What's the big deal?!

edit: yes, those are singular.

I'm currently using Primary: P20 & P17 and Secondary: P21 & 19   I was on the phone with another tech rep yesterday, and he agreed that the temperature my GPUs are currently at during idle and load are unusual. Generally people have a 50-60ºc in idle, and 70-80ºc max load not 70-80ºc idle, and 95-105ºc max load...   The tech said he was going to have a machine built with the same exac specs as my current computer and try to replicate the problem and fix it from there --however I don't see how this will work...   Edit: Also, Tristan, I used your thread before all this nonsense to try to see if any of this worked, and it does seem as if I can play my games, but I want to fix the temperatures on my GPU --I don't want to melt my hardware.

If everything is working now and you are just worried about the GPU temps then I say let them melt:) Dell will replace them when they do. I wouldnt worry about it. If its not working then read on.

Try this. Put 20 on the top cards 8-pin ONLY. Put 18 on the bottom card 8-pin ONLY. DO NOT use the remaning 6-pin on these two cables. Then run the remaning two 6-pin connections on both cards from the other cable you havn't used yet. I think it would be the one that P17 is on.

This picture is the way the chart says to do it and it is 100% wrong. Keep it like this picture but remove P19 and P21 connectors and use the remaning cable on those two instead.

When you start a game and the system just shuts off it means you are pulling more power then the PSU can handle.

You have ether:

1) Incorrect connections - know issue. Once you know this is not the case and you have nothing overclocked move on down the list. Keep in mind if you do not have it connected right it can make the issue look like its number 3. You need to make sure it is 100% correct.

2) Bad PSU - Checked with a tester

3) Inadequate PSU - Easy to check by using 1 card on a non-overclocked system. Make sure you do this with each card to rule out the GPU's as the issue. If it runs one card at a time but not two cards at a time....

4) Gremlins

Pray it's not number 4. That can be anything from a short,drivers,anything really and you will play hell trying to find it. Most of the time you would make sure everyhting is updated and then just start replacing parts untill you find it. If its a short in a cable then you're in real trouble. Dell can't find that. The system will have to be sent back to them or replaced. Good news is this one is very rare. Also, if Dell sends a tech to your house make him put a tester on the PSU.

Another thing you can check is what PSU you have. Dell sent me a 1kw once when it was supposed to be a 1.1kw. I only found out AFTER the tech installed it and left. The system would shut down when gaming so I had to pull a card untill they could ship me the correct one.  Pop it out and read the side. Can't hurt anything.

Also, confusing to me is my ATI Overdrive on my Catalyst showing me 0% activity; why?!

Also, confusing to me is my ATI Overdrive on my Catalyst showing me 0% activity; why?!

I noticed that on my 5870 also. When idle or running normal windowed apps in Windows, it reads zero and the  GPU clock is only 50% of full speed.

When I'm gaming, the GPU clock advances to full speed and Activity is usually around 90%. If you have 2 monitors, you watch the the CCC window while gaming full screen on the main monitor.

I think this is all normal. It's clocked down (when un-needed) to save power (and be able to keep the fan slower, etc.) The Activity gauge seems to be measuring 3D usage.

I see, kinda confusing, but I believe I understand; thank you. Anyways, are the drivers Alienware/Dell recommends really the only ones that will make useage out of CrossfireX? I updated to 10.9 originally and it prevented me from utilizing my CrossfireX. I havn't tried other drivers aside from that one and the 2 on Dell's site. Can anybody recommend a driver for alienware's dual 5970s?

Ugh, I'm not really having problems with games playing, but I'm down right frustrated about the whole GPU getting cooked; I can't seem to find the problem to this. My GPU idles at 75ºc and caps out at 96ºc now. And after looking at everyone else's reviews, it shouldn't be in the 90's. Most people say 50-60ºc idle and 75-85ºc load! Why is mine that much hotter? Singular they reach that heat too! :(

I don't think it's my rail connections anymore, because I havn't had any error's or random reboots for a while now. But for $7,500 I shouldn't be experiencing anything short of perfection. O_o

I don't think RAM can create an overheating issue.

That leaves the mobo or psu...

I did however recently get ONE bsod with the following description: Sm not used to checking all these things; heh...

Thank you for your feedback!