GTX680 4GB - ROM Image - Full Boot Screens, Full Clock Rates & PCI-E 2.0 - 5GT/s

cwright said:

I see the potential is there for every bit of performance I want, but there are two separate (I think) issues: Most games will average at 60fps or far beyond that, but with frequent dips down to 30-40fps, which makes for a very noticeable hiccup in gameplay. Second, after a few minutes, most games become unplayable at under 10fps.

Click to expand...

If this happens in one game it could be a memory leak. If it happens in all games, then yes, it sounds a lot like failing hardware or overheating. Could be someone had a bum card and instead of resolving it through a warranty claim, simply sold it off to you.

I noticed you checked the CPU temp but you should be checking the GPU temp.

I don't think memtest will show anything, but I won't discourage you from it because I think everyone should run this once anyway. To boot from a bootable disc, stick the disc in and then reboot. Hold down OPTION (ALT on a windows keyboard) at boot time and you will get the boot selector menu. This of course assumes you've flashed the ROM and can see boot screens. If you have not, boot into OS X and select the startup disc in system preferences.

ActionableMango said:

If this happens in one game it could be a memory leak. If it happens in all games, then yes, it sounds a lot like failing hardware or overheating. Could be someone had a bum card and instead of resolving it through a warranty claim, simply sold it off to you.

I noticed you checked the CPU temp but you should be checking the GPU temp.

I don't think memtest will show anything, but I won't discourage you from it because I think everyone should run this once anyway. To boot from a bootable disc, stick the disc in and then reboot. Hold down OPTION (ALT on a windows keyboard) at boot time and you will get the boot selector menu. This of course assumes you've flashed the ROM and can see boot screens. If you have not, boot into OS X and select the startup disc in system preferences.

Click to expand...

Thanks for the reply. Haven't tried every game yet, but it is happening on several of them - Bioshock Infinite, Metal Gear Rising, Far Cry 3 so far, for example. Just tried Far Cry 3 till it slowed down and eventually crashed - right after the temperature on the GPU read 45 degrees with a max temperature of 78 degrees Celsius. Don't know if that's normal or not.

Is there anything else that I should try before contacting the seller and attempting to return it?

What program are you guys using to monitor current clock and temperatures?

cwright said:

Is there anything else that I should try before contacting the seller and attempting to return it?

Click to expand...

Check the GPU temps.

Clean dust buildup off the heatsink and fans.

Make sure the fans are spinning.

ActionableMango said:

Check the GPU temps.

Clean dust buildup off the heatsink and fans.

Make sure the fans are spinning.

Click to expand...

Using CPUID HWMonitor for checking system info. All fans are spinning. The GPU temp on idle after first booting the computer is 32 degrees Celsius. After a game crashes or slows down, it reads about 45 degrees but also indicates a max temperature of 78 degrees. Is this a normal increase in temperature when under heavy load?

I did run the memtest just to be sure - no errors there.

The seller is willing to take the card back, but I want to identify the issue. I'd rather not return it, and if I have to want to make sure I don't run into the same problems with a new card.

cwright said:

Using CPUID HWMonitor for checking system info. All fans are spinning. The GPU temp on idle after first booting the computer is 32 degrees Celsius. After a game crashes or slows down, it reads about 45 degrees but also indicates a max temperature of 78 degrees. Is this a normal increase in temperature when under heavy load?

Click to expand...

78C is perfectly fine. I have no idea what's wrong then, sorry.

ActionableMango said:

78C is perfectly fine. I have no idea what's wrong then, sorry.

Click to expand...

I used the EVGA Precision app to limit power to the GPU to 80%, and left all other settings alone. I've only tried one game so far (Far Cry 3) but with that set, I haven't had any more crashes or slowdowns.

I'll test it more this weekend and see how stable it is. I guess no one else has had to limit power to the GPU to prevent these issues?

Ran into small issue. was using my friend's windows box trying to flash my zotac 680 card . Got the error attached. Any suggestions?


Thanks

Alex


Edit: think i found the solution.

Originally Posted by mcnallym View Post
Got my Zotac 680 today. Initially when trying to flash the card then refused with PCI Subsystem Mismatch.

Crossed Fingers, did the nvglash -4 -5 -6 name_of_rom.rom file and forced on.

Got it flashed and installed. Works great. Quick question - do i have to adjust frequency or just flash the rom posted in first post?

kalex said:

Got it flashed and installed. Works great. Quick question - do i have to adjust frequency or just flash the rom posted in first post?

Click to expand...

Adjust what frequency?

cwright said:

I used the EVGA Precision app to limit power to the GPU to 80%, and left all other settings alone. I've only tried one game so far (Far Cry 3) but with that set, I haven't had any more crashes or slowdowns.

I'll test it more this weekend and see how stable it is. I guess no one else has had to limit power to the GPU to prevent these issues?

Click to expand...

I just wanted to post an update here. I thought this had fixed the problem - it didn't. I hadn't had time to play with it much until this past weekend, but I'm still getting the same problem. Good news (I guess) is it seems to be it's not the graphics card that's the problem. I put my old card back in (MSI AMD 6870), turned settings down low enough to get a decent frame rate, and sure enough it still drops down to < 10 fps. Even quite a bit sooner than the GTX680 did.

Same as before, if I Alt-Tab to the desktop and then back to the game it will be OK for a short while before stalling again. Happens on multiple games. I've run FurMark GPU stress test on both cards and on both I see the frame rate dip there too - but only once, and afterwards it seems to fix itself indefinitely.

So I guess this is a software issue. I have latest drivers installed, on Windows 8. I think the next step is to start over with Windows 7 and try that. Any other ideas would be appreciated though!

cwright said:

Any other ideas would be appreciated though!

Click to expand...

I have no idea. I can only suggest general troubleshooting.

Run MEMTEST from boot; this should identify bad RAM. Bad RAM causes a plethora of problems, including some that wouldn't seem related to RAM at all. Be sure to run it from a boot source... running it from within an operating system prevents it from checking the RAM that the OS has reserved for itself.

Run the Apple Hardware Test.

If you have the ability, it would be good to start with a fresh install of the OS, fully updated, and with nothing installed but the game and the lastest video card drivers. This will eliminate software issues.

kalex said:

Ran into small issue. was using my friend's windows box trying to flash my zotac 680 card . Got the error attached. Any suggestions?


Thanks

Alex


Edit: think i found the solution.

Originally Posted by mcnallym View Post
Got my Zotac 680 today. Initially when trying to flash the card then refused with PCI Subsystem Mismatch.

Crossed Fingers, did the nvglash -4 -5 -6 name_of_rom.rom file and forced on.

Click to expand...

I'm getting this same error with an EVGA card. Can you, or someone, clarify how you fixed it? I'm not sure how to do the process in the last line of your post.

Thanks.

EDIT: I'm seeing this has been addressed earlier in the thread, so I'm back-reading. Sorry.

...Command line with bootflag trick from page 4 worked perfectly.

ActionableMango said:

I have no idea. I can only suggest general troubleshooting.

Run MEMTEST from boot; this should identify bad RAM. Bad RAM causes a plethora of problems, including some that wouldn't seem related to RAM at all. Be sure to run it from a boot source... running it from within an operating system prevents it from checking the RAM that the OS has reserved for itself.

Run the Apple Hardware Test.

If you have the ability, it would be good to start with a fresh install of the OS, fully updated, and with nothing installed but the game and the lastest video card drivers. This will eliminate software issues.

Click to expand...

I'll probably start looking for a solution on other forums since this is getting a bit off topic, as the problem has nothing to do with this card in particular. That said, a quick update:

Yesterday I wiped the drive and installed a new Windows 7 system and then nothing but graphics card drivers and Far Cry 3. I played it for several hours and it never had an issue - 60fps w/ Vsync and no drops. Last night I went ahead and installed several other games. Didn't run any of them - just wanted to get them loaded. Today I launched Far Cry 3 again and the same problem was back. In less than a minute it becomes slow and unplayable. Doesn't matter if you started playing right away or let it sit on the menus. It drops to < 10 fps and doesn't recover.

I uninstalled EVERYTHING else that I had loaded last night and rebooted, but the problem remains. I have no idea what could have been installed to cause this that I wasn't able to find. And now I know this happens on both Windows 8 and Windows 7, and on multiple graphics cards (AMD 6870 and the GTX 680).

This is so frustrating

cwright said:

This is so frustrating

Click to expand...

I feel like you've sufficiently eliminated software as a problem. And with the problem occurring on two different video cards (two different brands even), it does not appear to be a GPU problem.

At this point I can only assume it's a problem with the MP's hardware.

Did you do MEMTEST or AHT yet?
Do you have any other cards in there?

ActionableMango said:

I feel like you've sufficiently eliminated software as a problem. And with the problem occurring on two different video cards (two different brands even), it does not appear to be a GPU problem.

At this point I can only assume it's a problem with the MP's hardware.

Did you do MEMTEST or AHT yet?
Do you have any other cards in there?

Click to expand...

I ran MEMTEST the other day and it didn't find any errors. I don't have a Apple Hardware Test disc to try that. I was able to find a disc image download for a Mac Pro 3,1 AHT but it didn't work on my 5,1. Strangely enough it didn't give me an error message - just kicked the disk out and booted back to the Mac OS. I'll have to see if I can find a proper retail copy.

You don't think there are software issues? I don't understand why it would have been working fine all afternoon yesterday and only stopped working after I installed more software.

EDIT: The only other card installed is an Areca RAID card. No PC drivers installed for it - I only use it on the Mac side. I haven't had any issues with it and would doubt that it has anything to do with this issue.

cwright said:

I ran MEMTEST the other day and it didn't find any errors. I don't have a Apple Hardware Test disc to try that. I was able to find a disc image download for a Mac Pro 3,1 AHT but it didn't work on my 5,1. Strangely enough it didn't give me an error message - just kicked the disk out and booted back to the Mac OS. I'll have to see if I can find a proper retail copy.

You don't think there are software issues? I don't understand why it would have been working fine all afternoon yesterday and only stopped working after I installed more software.

Click to expand...

If you have had same issue in OSX and Windows you have ruled out software.

Maybe a dying HD?

MacVidCards said:

If you have had same issue in OSX and Windows you have ruled out software.

Maybe a dying HD?

Click to expand...

I haven't had any issues with it under OSX, but then again I don't run any games in OSX. If I weren't playing games on Windows I probably wouldn't notice any issues there either.

I would try Unigine Valley or Furmark in OSX.

cwright said:

I ran MEMTEST the other day and it didn't find any errors. I don't have a Apple Hardware Test disc to try that. I was able to find a disc image download for a Mac Pro 3,1 AHT but it didn't work on my 5,1. Strangely enough it didn't give me an error message - just kicked the disk out and booted back to the Mac OS. I'll have to see if I can find a proper retail copy.

You don't think there are software issues? I don't understand why it would have been working fine all afternoon yesterday and only stopped working after I installed more software.

EDIT: The only other card installed is an Areca RAID card. No PC drivers installed for it - I only use it on the Mac side. I haven't had any issues with it and would doubt that it has anything to do with this issue.

Click to expand...

Did you run MEMTEST from its own boot disk? I know it's more trouble to do, but running it after the OS loads is nearly useless because it cannot check for problems in areas that the OS has reserved for itself.

The AHT should have came with your mac either on Application Disk #2, or in a recovery partition on the original hard drive.

ActionableMango said:

78C is perfectly fine. I have no idea what's wrong then, sorry.

Click to expand...

Yeah, old Fermi 470/480 would top out over 90c. Ran fine there.

KFA2 GeForce GTX 680 EX OC 4GB on MACPRO 3.1

Hi any feedback about KFA2 GeForce GTX 680 EX OC 4GB on MacPro 3.1 quad

this is the card

http://www.kfa2.com/GTX680_4GEXOC.shtml

I will use the 2 x 6/6 pin cable that are connected to the ATI 4780 512MB

thanx

fpdesign said:

Hi any feedback about KFA2 GeForce GTX 680 EX OC 4GB on MacPro 3.1 quad

this is the card

http://www.kfa2.com/GTX680_4GEXOC.shtml

I will use the 2 x 6/6 pin cable that are connected to the ATI 4780 512MB

Click to expand...

Cables and power are fine. I don't think anyone has flashed that particular model yet, so you'd be the first to try that I'm aware of.

This is amazing, I cant thank you enough!

I've been pushing off upgrading my GPU because of ridiculous cost and crappy options. While searching around I came across the GTX680 Mac edition but I was turned off by the prices I was seeing... Then I found out that most GTX 680's could be flashed with the Mac ROM... WOW this is fantastic! I originally was debating about purchasing a Classified card with supplemental power supply in the 5.25" drive bay but I decided I did not like the idea of modifying the case and having a tons of wires everywhere, so I settles for the next best thing... EVGA 4GB FTW, I scored a used card on eBay for just a little over $200. The card appears to be in great shape and flashing the the 4GB ROM was a piece of cake!

After flashing the 4GB ROM, I adjusted the settings to match the default settings.


I wanted to load Nvidia Inspect to take a look but for some reason when I open the application, nothing loads. What am I doing wrong?



Anyway, it boots into OSX just fine and I have the EFI loading screen!!!



EDIT: Thought I'd post up my benchmarks.

Cinebench
OpenGL 53.71 fps
CPU 834 cb

GeekBench 3
Single 2902
Multicore 16581

Zotac GTX 680 4GB issues….

I flashed my card using the exact instructions from Dr. Stealth's post. Saved the original ROM from my Zotac GTX 680 4GB and used Kepler tweaker to match setting to the GTX680-4GB-ROM.rom, saved the modified EFI rom and flashed using command prompt with -5 -6 arguments. I am getting constant square, green artifacts on screen with freezing GUI. Console is showing this error:

kernel: NVDA(OpenGL) Channel Timeout!
kernel: NVDA(OpenGL) Channel exception! exception type = 0xd = Graphics Engine Error (GR Error 5)

It causes me to have to do hard restarts, which corrupted my boot drive and time machine backup! -Luckily I used MacDrive in Bootcamp to recover data.

I was having these issues in Mavericks 10.9.4, i read there's issues there so I reinstalled the OS to get it back to 10.9 and then updated to 10.9.2. Still had the same issues. I installed the Nvidia Web Drivers for 10.9.2, I thought it was okay for a day, then Photoshop CS6 started crashing the computer.

All of those hard restarts from the NVDA(OpenGL) Channel Timeouts corrupted my SSD boot drive. So I had to wipe if from Bootcamp, because Disk Utility would not unmount it for a repartition when booting from a USB installer.

I took that opportunity to reinstall Mountain Lion 10.8.5. I flashed it back to the original PC ROM and installed Web Driver 313.01.03f01 and it seemed okay, so I reflashed the 4GB EFI ROM to it and I got those artifacts, so I PRAM reset back to OS X Default Drivers. I cannot tell if I have a bad card, bad rom, bad logic board, bad RAM? I know the card is not overheating because I run Magican to monitor temps and the GPU rarely gets to 67° when using Parallels. Running Unigen Valley has never once crashed the card. I ran the GPU-Z render test and it doesn't crash in Bootcamp. However, I do occasionally get an error that the Nvidia Display Driver stopped responding when booting Win 7 x64 SP1, but it hasn't ever crash Windows, just OS X.

Tech Tools VRAM test shows it passes. I have 2009 Mac Pro, that I flashed back to the 4,1 firmware trying to troubleshoot this.

Anyone have any thoughts on this or experienced the same thing? Also if someone would like to share their ROM being used in their healthy and functioning Zotac GTX 680 4GB card, that would be much appreciated too. I really want to get back up to Mavericks 10.9.2 because all my FCPX projects are now in the FCP 10.1.x format and can't be opened in 10.8.5.

My system:
2009 Mac Pro 4,1
2.66GHz Xeon W3520
256 SSD in lower optical bay
1 TB, 500GB and 200GB in HDD bays 1 - 3.
4 port Inateck USB 3.0 PCIe card in slot 4 being powered by SATA port 4
Zotac GTX 680 4GB (2 6pin PCIe power) in PCIe Slot 1
32 GB (16x2) ECC RAM