Hard drive issue after installing new graphics card

ssp212

Distinguished
Mar 5, 2009
38
0
18,530
Hello,

I have just bought a new case (Antec Three Hundred Two) and a new graphics card (ASUS Strix GeForce GTX 970). I have transferred the rest of the machine from my old case:

Windows 7 Home Premium 64 bit OS
Dell 2408WFP monitor
750W Be Quiet Dark Power Pro power supply
6 GB Corsair DDR3 RAM
Gigabyte EX58-UD4 motherboard
Intel Core i7 920 D0 processor
1 TB WD Black (WD1003FZEX)
Samsung SH-118BB DVD-ROM

My old graphics card was a SAPPHIRE HD 4890 and my old case was an Antec Three Hundred.

Ever since setting this up, I have had intermittent freezes in Windows (the hard drive appears to stop responding). The system recovers after a short time and carries on as if nothing had happened. This does not occur in Safe Mode, but does seem to occur using the most basic Diagnostic Startup. I can see event "the device, \Device\Ide\iaStor0, did not respond within the timeout period" in Event Viewer after each occurrence but no other error messages, blue screens, etc.

About a year and a half ago I bought an MSI GeForce GTX 770 GAMING and had similar (worse) problems. At the time I was having difficulties with my hard drive and put it down to that (I ended up sending the 770 back, putting the 4890 back in and replacing the hard drive and optical drive). I opened a thread about it here: http://www.tomshardware.co.uk/answers/id-1760179/hard-drive-issue-installing-graphics-card.html

Between then and now the system has functioned normally with no problems whatsoever.

I have tried:

- Performing a SMART drive quick self-test (passed and assumed OK as no previous problems with this drive).
- Removing all trace of AMD / ATI drivers including editing left over registry entries.
- Altering power management options and adding registry entries to disable Link Power Management.
- Using MSAHCI drivers instead of iaStor.
- Using different cables (data and power) for the hard drive.
- Using different sockets on the PSU for the hard drive.
- Using the Gigabyte GSATA sockets instead of the Intel SATA ports on my motherboard.
- Switching the drive to IDE mode in the BIOS.
- Disabling NVIDIA HDMI Audio

I haven't tried:

- Moving the new graphics card to a different slot.
- Reinstalling Windows.
- Testing RAM (I assume this is OK as there were no problems previously).

I'm at an absolute loss and I don't want to give up like I did last time! Does anyone have any suggestions? Is there some fundamental incompatibility that I am missing? I realise there are parts of my system that are getting a bit old, but I assume the motherboard and PSU are still OK.
 
Actually, while the Dark power pro 850w and up units are Tier 1 units, the DPP 750w and lower units are made by a different OEM and are Tier 3 units, which are not generally recommended for overclocking OR for use with high end gaming cards. You might want to consider a higher tiered PSU. I don't think, in a preliminary off hand way, that is likely your main issue though, but it could be.

PSU Tier list: http://www.tomshardware.com/forum/id-1804779/power-supply-unit-tier-list.html

I'd recommend you start by running Seatools for Windows on all connected drives. The SMART tests are very limited. The Seatools utility is more comprehensive. Run the short DST and long generic. Drive brand is irrelevant.

Seatools for Windows: http://www.seagate.com/support/downloads/item/seatools-win-master/


I would then perform a CLEAN installation of the GPU drivers, regardless that you've taken some similar steps previously, as outlined at the following link:

http://www.tomshardware.com/faq/id-2492424/laymans-simple-guide-solving-graphics-card-issues.html


You definitely want to use the Intel SATA headers to connect the drive on your motherboard, so if you're still connected to the GSATA headers, switch them.

Your graphics card should be in the #1 slot, closest to the CPU. If all else fails, we can try moving it to another slot later, but if the slot is the problem you may want to replace the motherboard as secondary slots on many boards do not offer the same performance as the primary slot. Some boards do, some don't.

If you've come this far with no results, I'd recommend running Memtest86 for a full seven passes. The number of passes can be set prior to beginning the test.

Memtest86: http://www.memtest86.com/download.htm



Running an SFC/Scannow might not hurt either:

http://www.eightforums.com/tutorials/3047-sfc-scannow-command-run-windows-8-a.html
 

ssp212

Distinguished
Mar 5, 2009
38
0
18,530
Hello,

I have followed the clean driver installation guide, used DDU and reinstalled the latest drivers. I have reseated the card, checked the cables and plugged the PSU directly into the wall. The card is in the nearest slot to the CPU and I am using the Intel SATA headers again (I have reset most of the options I mentioned in my first post once I determined they made no difference).

I have also updated the Intel Rapid Storage Technology driver to version 11.7.0.1013 (I understand this is the most recent version to support the ICH10R chipset). Since doing this the message in event viewer appears to have changed to "Reset to device, \Device\RaidPort0, was issued" but the symptoms are the same.

I'm currently running MemTest86 with no errors reported so far (I wasn't sure how to set the number of passes, so I am just letting it run at this point).

Do you have any other suggestions for if and when MemTest86 completes?

I would very much appreciate your advice!
 
Check your BIOS version. If it's older than version F11T, then I'd suggest you update to version F11T located here:

http://www.gigabyte.com/products/product-page.aspx?pid=2988#bios

You don't need to update every version, updating to the the most recent version will include all previous updates as well. There seem to be many different version updates so it's likely there were some issues with hardware compatibility.
 
Solution
Since it's been in beta since 2011, it's unlikely there were issues with it as nothing superceded it and it has not been changed since. If there are any issues you can probably try the one just below that or go back to F10 I would think. It should be ok. I'll double check to see if there are any reported issues with it.

 
I don't see any reported issues with it, but as always, be aware that flashing the BIOS, CAN be, a risky endeavor and that I take no responsibility for any issues that result from doing so. That being said, it should be fine, I flash BIOS all the time with early releases, which they don't even bother calling beta anymore most the time. Maybe try the one below that one first, see if it clears up your issue and then try the newest one if it does not.

On the other hand, there must have been some further hardware compatibility issues that the F11o firmware did not fix or there wouldn't have been another version released.

You're sure your board is UD4 and not UD4P right?
 

ssp212

Distinguished
Mar 5, 2009
38
0
18,530
Hello,

Thanks for sticking with me! I have updated the BIOS to F11o and now when I try going into Windows normally, the machine resets itself (the black screen with the Windows logo appears, then the mouse pointer appears on a black background and then the machine resets). I can still get into safe mode with no problems. I will give F11T a try!
 
Ok, if that exibits the same behavior you may need to go back to the the F10 version. Computers can be fickle sometimes. You might be well served to also try running the system with only a single memory module installed. If the problem still occurs, switch to another module. Make sure the module is installed in the slot designated in the motherboard manual for single module operation.
 

ssp212

Distinguished
Mar 5, 2009
38
0
18,530
I have now updated to F11T and things seem to be working normally (the hard drive issue has not happened yet either, but I want to give it some time before I celebrate too much).

One thing I did with F11T, which I didn't do with F11o was to load Windows without changing any settings in BIOS at all. I verified that it worked and then went back and changed some settings, e.g. enabling AHCI, etc. I noticed that when I enable C3 / C6 / C7 State Support, F11T crashes and restarts on loading Windows as well. I used to have C3 / C6 / C7 State Support enabled in F10 with no problems (before the graphics card upgrade), but now I'm wondering if this was the cause of the hard drive issue in F10 in the first place. I'm not 100% sure of the purpose of this setting, but the motherboard manual suggested it could decrease power consumption and would only be available if the CPU supported it.

Does this make sense / sound feasible? If so, would you suggest going back to F10 or F11o and disabling this setting, rather than staying with F11T?
 
Your power supply may not support some of the C states. Haswell chips that support C6 and C7 states need Haswell complaint power supplies with good DC to DC conversion support. All of the current BeQuiet power supplies are Haswell compliant for the C6 and C7 states.

Additionally, I don't think the i7-920 supports the C6 and C7 states anyhow, and could actually be causing an issue. I know it supports C1 and C2 but I think there issues with using that chip with C6 and C7, plus, it seems on many boards there are also issues of conflict with the turbo boost and those two C states that doesn't allow having both on at the same time on some chip/board combinations. I'd leave them off.

If you have anything overclocked, it's pointless anyhow.
 

ssp212

Distinguished
Mar 5, 2009
38
0
18,530
Hello,

Just wanted to say that I tried going back to F10 and changing that setting, but again couldn't get into Windows. I'm back on F11T with the C3 / C6 / C7 State Support disabled and all other settings as I want them. I've only had time to use the machine for a few hours over the weekend, but everything seems fine and I've been able to briefly have a go at a few games and enjoy the new card!

Thanks for your help. It was much appreciated.