22 replies to this topic

[Roadbuster]

After upgrading from my previous GFX card (MSI Radeon 5850) to a Nvidia 660ti I'm experiencing a problem with MWO.
Sometimes my screen goes black for a few seconds, then comes back and I'm running with ~10fps. After some time and a few Alt-Tabs the game would run normal again but sometimes without sound.

A few times I still saw the message that the display driver was restored after an error.
So I started looking for the source of that problem.

What I did so far:
- tried different drivers: newest, WHQL certified, from GFX Card driver disc
- tried different game settings
- upgraded PSU from a 550W Enermax Modu82+ to a 860W Corsair AX860i (although I don't think power was the problem, Cyber Power USV showing ~170W idle and ~300W under load)
- new installation of Windows 7 Ultimate 64bit

I thought that maybe some driver was left somewhere from the ATI card in the Windows installation but then the same error occured again after a fresh install of Windows.

I'm trying to play with all settings on medium, no AA and activated Vsync now.
Also changed driver settings to overwrite the game settings, disabling everything I could except Vsync.

Still got more testing to do to see if it's working with these settings, but so far the error didn't occur again. The game doesn't look as good as it could though.


Someone else got this problem or maybe a sollution?

[Coastal0]

This sounds ********, but how's your heat?

I currently have a 660Ti, almost brand new. This has been working wonders for me.

So, why am I posting in this thread, you ask?

The card I changed *FROM* was a Radeon 6970. I was getting really bad FPS, and I had a similar error message to you quite regularly, and experienced similar issues. (Also tested with fresh windows install, driver wipes, etc.)

My GPU was running ~97 degrees Celcius.

My new one, the 660Ti, is running quite a lot lower than that.

[Flapdrol]

Graphics cards can have defects, sounds like yours is unstable, driver crashes and recovers but starts running at "safe" (really slow) clocks.

Maybe try another driver, test a few other games. If you still get the problem return the card to where you bought it and get a new one.

Edited by Flapdrol, 14 February 2013 - 03:37 AM.

[K4mil]

artifacts?

[Silpher K]

if POSSIBLE, (i know it isn't always possible for some people), try installing your 660ti in another computer and see if the problem follows the card. then you know if it's other factors or not.

[Eyecon69]

Hey I know this is going to sound absolutely bonkers but double check all your IRQ settings in your Bios... I had turned on my serial port to update some firm ware, instead of turning it off when i was done, I went to the value below which was a static irq and it wreaked havok on my pc.. I hadn't noticed for the longest time until after I updated my cards...

its worth a shot.

Edit: also what is the brand name on the card. Some seem to have better QA than others.

Edited by Eyecon69, 14 February 2013 - 11:52 AM.

[Catamount]

well Windows has IRQ errors (specific BSODs that occur); this honestly sounds like a faulty card, which I just ran into myself a couple weeks ago (7970, replacement works beautifully).

Try other games and software and see what happens. Get the new 3dmark, play something that chews systems up like Kerbal Space Program, run a GPU stress tester, just see how the card behaves in general.

[Roadbuster]

Thx for your advice everyone.

The card is a Palit Geforce GTX 660Ti Jetstream.
Temp of the card usually is around 35-40°C idle and ~70-75°C under full load with furmark, measured with GPU-Meter. No errors, artefacts or anything else during testing.
No crashes, errors high temps anywhere. No dust on any coolers (CPU is watercooled anyway)

Checked IRQs, no problems there.
No artifacts or flickering or something like that.

Going to try more games to see if there's a problem anywhere else, but so far this only happened in MWO.

Also, since I changed the driver settings and set all in-game graphic settings to medium, there were no problems.
Didn't get a single error in ~30 games yesterday compared to 1-3 errors every 5 games before.
I used to run the game at medium settings with my previous card too and never had a problem.

I guess it's either a problem of the MWO graphic settings or the GPU driver or some combination of both that makes problems.

What are your game settings?

Edited by Roadbuster, 14 February 2013 - 11:45 PM.

[Catamount]

Roadbuster, if you can even run Furmark, then that confirms it's most definitely not you card. That software is so stressful I hesitate to even recommend it, because I've seen it brick GPUs before.

This is definitely either MWO or a driver problem. Try running MWO at higher settings and see if the problem is fixed with the new driver.

If it isn't, you can do a complete reinstall of MWO, and then if that doesn't work, then wipe the drivers out completely (as in uninstall them, and then get rid of all folders and registry entries; there should be guides on that around or software that does it) and reinstall, which I had to do with my AMD drivers once not that long ago. Driver Sweeper used to be the way to do this, but iirc, it's under a new company and is spyware now. An older version would still be fine though.

Edited by Catamount, 16 February 2013 - 05:42 AM.

[Eyecon69]

Roadbuster, on 14 February 2013 - 11:40 PM, said:

Temp of the card usually is around 35-40°C idle and ~70-75°C under full load with furmark, measured with GPU-Meter.

My 8800 Ultras were cooking at over 90 degrees... Apparently this is considered pretty safe and normal for THAT card... The chrome on the mounting brackets actually turned blue LOL!!!

Back to your problem... I run triple monitors on medium resolution. I have an E8500 Core 2 Duo so its not worth it running to much higher. I CAN run at a higher resolution but but I'm running internet off my phone right now and I notice when I'm at the higher resolution I can't watching other place after very well... And I'm a green bean so it helps lol

I would wonder you say other games work fine are any of these on the Cry engine? If its only MWO and other cry games work fine it could be something in the launcher.

Also depending on where you bought the card you might be able to talk to the shop guys and they could test it for you. I've found sometimes buying locally can have its rewards. Also I'm sure some one you know some where has a lesser card and might be thinking of upgrading you could "Offer" to let them test yours to see if there is a super awesome performance gain.. its a win win... or lose lose :S

[Roadbuster]

Catamount, on 15 February 2013 - 09:35 AM, said:

Roadbuster, if you can even run Furmark, then that confirms it's most definitely not you card. That software is so stressful I hesitate to even recommend it, because I've seen it brick GPUs before.

This is definitely either MWO or a driver problem. Try running MWO at higher settings and see if the problem is fixed with the new driver.

If it isn't, you can do a completely reinstall of MWO, and then if that doesn't work, then wipe the drivers out completely (as in uninstall them, and then get rid of all folders and registry entries; there should be guides on that around or software that does it) and reinstall, which I had to do with my AMD drivers once not that long ago. Driver Sweeper used to be the way to do this, but iirc, it's under a new company and is spyware now. On older version would still be fine though.

You are right. Furmark is kinda hardcore for a GFX card, but at least you will really know if your system is stable under heavy 3D computing load.

I'm going to try running the game at high settings with Nvidia drivers overwriting the game settings next. With medium settings there was not a single error in ~80 games now.

The game is running on a fresh Windows installation now.

Eyecon69, on 15 February 2013 - 04:34 PM, said:

I would wonder you say other games work fine are any of these on the Cry engine? If its only MWO and other cry games work fine it could be something in the launcher.

Also depending on where you bought the card you might be able to talk to the shop guys and they could test it for you. I've found sometimes buying locally can have its rewards.

I currently don't run any other game with the Cry engine. But it could be possible that it's a problem of GPU-diver-engine combination.

And you're right with buying components from a local dealer. I do that too.

[Gonzo007]

hm, i with same GPU as original poster, not played few months, and after launched game today GPU go up to 68c and then boom at some moment screen just froze and game in manager "not responding" (when i played last time everything was normal)

[00 Zedd]

I have the Asus 660ti, everything runs buttery smooth, with all graphic settings on full. Rarley does MWO freeze and crash to desktop; But I think everyone has that problem.

Sounds like you have a defective card, or maybe it does'nt play nice with the rest of your computer.

[Gonzo007]

00 Zedd, on 18 February 2013 - 06:58 AM, said:

I have the Asus 660ti, everything runs buttery smooth, with all graphic settings on full. Rarley does MWO freeze and crash to desktop; But I think everyone has that problem.

Sounds like you have a defective card, or maybe it does'nt play nice with the rest of your computer.

as i said, i already played cbt and obt, everything was fine. I play lots of other games, never had any freezes

[HRR Mary]

Using the MSI 660 card TwinFrozer, and I have been experiencing the same issues.

The black screen seems to indicate that the card powers down (fan stopping, MSI afterburner showing drop in power usage), Alt-Tabbing for a few seconds to desktop resumes full FPS.

On my card, I think I have pinpointed the issue, being a faulty heat probe (car registering 60°C max at full charge, which strikes me as odd).

OP, if I were youI would follow the advices of returning the card for an exchange (which is exactly what I'll be doing this week).

[Flapdrol]

Catamount, on 15 February 2013 - 09:35 AM, said:

Roadbuster, if you can even run Furmark, then that confirms it's most definitely not you card. That software is so stressful I hesitate to even recommend it, because I've seen it brick GPUs before.

This is definitely either MWO or a driver problem. Try running MWO at higher settings and see if the problem is fixed with the new driver.

If it isn't, you can do a complete reinstall of MWO, and then if that doesn't work, then wipe the drivers out completely (as in uninstall them, and then get rid of all folders and registry entries; there should be guides on that around or software that does it) and reinstall, which I had to do with my AMD drivers once not that long ago. Driver Sweeper used to be the way to do this, but iirc, it's under a new company and is spyware now. An older version would still be fine though.

But furmark isn't a good stability test. You can have a card that's stable for hours in furmark crash in 3dmark.

[Grrzoot]

It is most likely a bad card. What nvidia doesn't tell people is that 660 ti's are actually 680 gtx that as a bad portion of silicon. They had a huge mess with the 680's when they first came out and had a loss of 75% or something on the production run, so what they did was disable certain areas of the card then rebrand them as a lower spec card, which is why on paper the 660ti is better than the 660. Although most of these cards run just fine with the bad portions disabled (think those amd 4 cpu chips that only ran three cpu's because one was bad) sometimes the boards have other quirks or issues. I would just return/exchange now before it really cooks itself.

[Werewolf486 ScorpS]

Download ATIMAN uninstaller and run that....

It is the best driver removal tool for radeon cards. Then download the Palit thundermaster and boost the voltage up a little at a time. The 550's had issues being under voltage and that was the cure. It will also monitor everything you need, and DO NOT run any other software to monitor or adjust while using this.

If nothing here fixes it then you should start ticket with the manufacturer.

[Glucose]

If you haven't already figured it out -- this is likely not a bad video card, as many people are experiencing their driver reinitializing since the Feb 5th patch.

[Atlas32]

Roadbuster, on 14 February 2013 - 11:40 PM, said:

Thx for your advice everyone.

The card is a Palit Geforce GTX 660Ti Jetstream.
Temp of the card usually is around 35-40°C idle and ~70-75°C under full load with furmark, measured with GPU-Meter. No errors, artefacts or anything else during testing.
No crashes, errors high temps anywhere. No dust on any coolers (CPU is watercooled anyway)

Checked IRQs, no problems there.
No artifacts or flickering or something like that.

Going to try more games to see if there's a problem anywhere else, but so far this only happened in MWO.

Also, since I changed the driver settings and set all in-game graphic settings to medium, there were no problems.
Didn't get a single error in ~30 games yesterday compared to 1-3 errors every 5 games before.
I used to run the game at medium settings with my previous card too and never had a problem.

I guess it's either a problem of the MWO graphic settings or the GPU driver or some combination of both that makes problems.

What are your game settings?

I have a new 660Ti that idles at about 29c and while playing MWO, I only break 35c...barely. You're seems to be running much hotter than mine