Asus HD 7970 ROG Matrix Platinum crashing within 10 minutes

[alex_stanhope]

Hi there,

I bought an Asus HD 7970 ROG Matrix Platinum last week and I'm having a terrible experience with it. It reliably crashes within 10 minutes when playing Battlefield 3. I'm using it with an Asus motherboard (P8Z77 WS) and a 650W FSP Aurum Gold PSU, which should provide plenty of power.

I've tried several different drives (13.9 and 13.11 beta), removed the CPU overclock (just in case) and even put it in another PCIEx16 slot, all with the same result. I have not overclocked the graphics card at all. I documented the crash reports and they don't seem to be temperature/overheating related.

11/10/2013
23:00
DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG: Device hung due to badly formed commands.. GPU: "AMD Radeon HD 7900 Series", Driver: 13.9 (13.152-130830a-163178C-ATI)
48.0 degrees

12/10/2013
07:45
DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG: Device hung due to badly formed commands.. GPU: "AMD Radeon HD 7900 Series", Driver: 13.9 (13.152-130830a-163178C-ATI)
43.0 degrees

DirectX function "device->CreateBuffer(&dxDesc, &resourceData, &buffer->m_buffer)" failed with DXGI_ERROR_DEVICE_REMOVED: Hardware device removed.. GPU: "AMD Radeon HD 7900 Series", Driver: 13.9 (13.152-130830a-163178C-ATI). This error is usually caused by the graphics driver crashing; try installing the latest drivers. Also, make sure you have a supported graphics card with at lesat 512MB.
53.0 degrees

14/10/2013
07:10
DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG: Device hung due to badly formed commands.. GPU: "AMD Radeon HD 7900 Series", Driver: 13.9 (13.152-130830a-163178C-ATI)
75.0 degrees

07:34
DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG: Device hung due to badly formed commands.. GPU: "AMD Radeon HD 7900 Series", Driver: 13.11 (13.20.16-130926a-1630660E-ATI)
72.0 degrees

21:09
DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG: Device hung due to badly formed commands.. GPU: "AMD Radeon HD 7900 Series", Driver: 13.11 (13.20.16-130926a-1630660E-ATI)
77.0 degrees

15/10/2013
DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG: Device hung due to badly formed commands.. GPU: "AMD Radeon HD 7900 Series", Driver: 13.11 (13.20.16-130926a-1630660E-ATI)
66.0 degrees

Unless you can suggest something to make it work reliably (Graphics Card BIOS flash?) I'd like to return it please.

Cheers, Alex

[Mark@SCAN]

Hi,

There is a BIOS/Firmware update for this card, please see below link:

http://www.asus.com/Graphics_Cards/M...P3GD5/#support

If this does not resolve your problem please submit an RMA request via the following link:

https://secure.scan.co.uk/aspnet/sup...px?QueryType=R

Regards

[alex_stanhope]

Thanks Mark,

I installed the firmware update, but the card is still crashing reliably. I appreciate your help. My query number is OLQ***** for the RMA.

Cheers, Alex

[Mick@SCAN]

alex_stanhope,

An RMA has been raised as requested and instructions emailed to you directly.

Kind regards,

[alex_stanhope]

I'm sorry to say this problem has not been resolved. I returned the card to Scan where it was tested. The service engineer, while very friendly, was not able to reproduce the problem. I paid for the card to be returned next day and for the engineer's time to test it, and the card was shipped back to me a couple of days later. Unfortunately, the problem is continuing and seems to happen more quickly than before.

I couldn't get much information on what tests were run at Scan HQ but only that the card didn't crash "running Crysis 3 for 3 hours". I don't know how that test was performed, but the crashing only seems to affect the card at very high load. The fans are working well and the card's temperature doesn't go above 70-75 degrees, but I suspect there's some other (unmonitored) on-board component that's overheating or failing. When I heard of the difficulties replicating the problem I tried to contribute more information. The service engineer didn't (or couldn't) reply to my suggestions. The RMA form, which seems to be my only way of contacting or feeding information to the engineer, seems to have a 24 hour delay on it. I imagine maybe Scan get a lot of SPAM, but when I was trying to resolve a problem my not having a way to contact the person testing the card made the whole process very difficult and frustrating. Would it be better next time if I conduct the whole dialogue through this forum?

I did wonder if my UPS was a problem, but having removed that from the equation, I'm still getting the crashing. It's got so bad now that the following games are unplayable:

Battlefield 3
Far Cry 3 Blood Dragon
Grid 2

I'm not sure what to do. I'm in the horrible position of having paid for an unusable card that crashes in every gaming session, but the intermittency of the problem makes getting it sorted out a painful campaign. If I return the card to Scan again I risk meeting the same null response on testing and paying another £30, not the mention all the time it takes to organise and execute an RMA. Asus don't publish a diagnostic so I've got to find another way of proving the card to be faulty.

I have patched the GPU firmware/BIOS with the latest manufacturer update and installed the latest drivers:

Driver 13.11beta9v2 (13.250.18.0)

On balance this has been an awful experience. I don't blame Scan for the faulty card - I love Scan and there's no way you could have known it was faulty when supplying it - but the failed RMA return has been disappointing. I really hope you can suggest some way that I can get this mess sorted out. I'm desperate.

[Mark@SCAN]

Sorry to hear this,

As I am sure you appreciate intermittent problems are the hardest to replicate, have/can you test the card in another machine to see if the problem follows the card?

We can test in BF3 however Crysis 3 is generally more of a stress on the card which is why it was used. Can you consistently reproduce the problem in BF3 if so after how long roughly does it happen?

Regards

[alex_stanhope]

Thanks for being so responsive,

I can consistently reproduce the BF3 crashing problem. In multiplayer (Conquest Domination 48), it always crashes within about 10 minutes.

I can't test in another machine because I've not got anything else that's both Windows based and will take a 3U PCI-E graphics card.

Do you have Far Cry 3 Blood Dragon? It crashes incredibly reliably in that within 60 seconds. I'd even buy Scan a copy if it meant we could prove there is a problem!

[Mark@SCAN]

Hi,

I'm pretty sure we have FC3BD and I know 100% we have BF3. I think it's best we get the card back again for testing, if the fault is confirmed we can refund the previous NFF charges and get the card replaced for you.

Please submit a new RMA request via the following link:

https://secure.scan.co.uk/aspnet/sup...px?QueryType=R you can link to this thread if you wish.

Regards

[MrRockliffe]

You should try uninstalling everything to do with graphics cards. I think I used a programme called driver fusion. Install it and then boot to safe mode and remove everything after having uninstalled the programmes (even direct x). I know this sounds stupid, but make sure the card is all the way in, and secured both by the pci clip if you have one and the screws. Then install everything again using the latest beta drivers and catalyst control. I had the same problem with a 7850 but did the things I listed and then it worked. You could also try installing Linux and running a game in that?

[alex_stanhope]

Thanks Mark, have requested the RMA (query number OLQ1216724) and I'm just waiting on approval.

Thanks MrRockliffe, I've done a complete reinstall of the AMD driver from scratch and it didn't make a difference. I think you're right to highlight the importance of properly installing the card properly. I've had poor connectors in the past, but I was very careful to check this installation (tried multiple PCI slots as you know).

[alex_stanhope]

I thought it might be helpful to post up the exact settings I've been using an experienced problems.

Far Cry 3 Blood Dragon used to be unplayable too (crashing within 30 seconds) but the latest beta drivers seem have fixed that. Battlefield 3 (multiplayer) and Grid 2 are still the worst offenders.

17:51 17/11/2013
Card temperature 79.0
Grid 2 crash, black screen of death
Driver 13.11beta9v2 (13.250.18.0)

19:23 18/11/2013
BF3 crash, produced error message, closed program
DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG: Device hung due to badly formed commands.. GPU: "AMD Radeon HD 7900 Series", Driver: 13.9 (13.25.18-131107a-164864-ATI)
Driver 13.11beta9v2 (13.250.18.0)

19:38 18/11/2013 (GPU monitoring open)
BF3 crash, produced error message, closed program
13.25.18-131107a-164864-ATI
Driver 13.11beta9v2 (13.250.18.0)

19:41 18/11/2013 (GPU monitoring open)
BF3 crash, produced error message, closed program
13.25.18-131107a-164864-ATI

19:43 18/11/2013
BF3 crash, produced error message, closed program
13.25.18-131107a-164864-ATI

21:56 18/11/2013
Grid 2 crash, black screen of death, pointer disappeared from desktop

21:59 18/11/2013
FC3BD crash while starting up, black window stalled on desktop

It's hard to generalise but broadly:
* BF3 plays ok for about 10 minutes, then starts crashing with the DirectX error above. Every time I reload it after that point it then crashes more quickly, usually just a couple of minutes.
* Grid 2 starts flickering black polygons before eventually giving up all together and freezing.
In neither case does the temperature on the card go too high. It's never over 79 degrees.

I've tried reinstalling the drivers, updating DirectX, even underclocking the card and nothing addresses these issues.

My setup:
1. Single screen, 1680x1050 (DVI-D to DVI-D)
2. Three screen Eyefinity (left-to-right), 1280x1024 (display port to VGA), 1680x1050 (DVI-D to DVI-D) running at 1280x1024, 1280x1024 (DVI-I to VGA using supplied adaptor)

I hope that's helpful.

[MrRockliffe]

Try only using one monitor. And create a new partition on your hdd and install windows to it and a game. Install necessary drivers and try again. If it works, copy over your stuff and delete the old partition

[sqwerty]

When you installed the game/s did the direct x install with it? I have seen similar issues that were fixed once the dxsetup from the game directory was run manually.

[alex_stanhope]

Thanks for the helpful suggestions. The card has been returned to scan and a fault found.

Mark@Scan, I've received a refund and as my original card (HD7970) is now end-of-life, I've ordered a replacement Asus GTX770. I haven't received a refund for the NFF charges (£20+VAT) from the first RMA where the fault wasn't found and my email from two days ago has gone unanswered. I don't mind if Scan decide not to refund the NFF charges but I'd just like clarity on what's happening please.

[Mick@SCAN]

alex_stanhope,

I can confirm that both your graphics card and testing charges were refunded on the 28th of November, please allow around 3-5 working days for your bank to process the transaction.

Kind regards,

[alex_stanhope]

Just want to say a massive thank you for sorting this out. I've bought a GTX770 (from Scan of course) and it's working really well.