Major Issue MSI 560 TI 448

[Lazermonkey]

Hey everybody,

I have the MSI GTX 560 Ti 448 core OC/PE

I am having a major problem with it.

Today I was trying to watch a Youtube video and noticed major stuttering.

Then I tried to load up Far Cry 2 and it crashed.

I also tried a couple of other games with the same results.

I noticed Afterburner is only showing I am using less than 100Mb of memory when I ran an OCCT test.

Is the memory fried? Something else?

This sucks as it is my only card and my board doesn't have onboard video.

I called MSI and got an RMA set up.

TIA for any and all input

The following pic is while the card was under a load.

 

[erocker]

Start by uninstalling all GPU overclocking software/utilities.

Make sure your drivers are up to date.

Check temperatures.

A clean uninstall an reinstall of your video drivers.

 

[Lazermonkey]

Start by uninstalling all GPU overclocking software/utilities.

Make sure your drivers are up to date.

Check temperatures.

A clean uninstall an reinstall of your video drivers.

I did all of that.

The temps on this card never got about 63c and that was during a benchmark.

The card is clean, no dust build up etc. as I keep my stuff cleaned up.

During gaming the card would average about 53-55c max.

I think it is done for? I can browse the internet etc. just anything requiring more power than that and it fails.

Will MSI send me another 448? I know they were limited edition cards. If not, will they send me a 570, possibly?

 

[Crap Daddy]

I would add to this to try and reseat the card. It's also a good opportunity to blow the dust off.

 

[Lazermonkey]

I would add to this to try and reseat the card. It's also a good opportunity to blow the dust off.

I already reseat the card.

Also, no dust. The card looks brand new and is less than a year old.

Running OCCT GPU test the card is only using 86mb of memory max.

 

[Crap Daddy]

Can you test the card with another PSU? Unigine Valley is a good test.

 

[Lazermonkey]

Can you test the card with another PSU? Unigine Valley is a good test.

I don't have another PSU.

I will try the Valley test.

Here is a test from the other day.

Here are the results from just now.

 

[Crap Daddy]

I don't have another PSU.

I will try the Valley test.

I can't think of anything that you might try except the PSU.

 

[Lazermonkey]

Any other suggestions?

 

[McSteel]

What does the initial GPU-Z screen say? Perhaps your PCI-E link speed has dropped somehow..? The GPU is not being fed information properly, this is why your VRAM usage is low, and why the GPU is so cool - it's vastly underutilized.

Perhaps you could flash a new motherboard BIOS, and check it's settings to verify PEG link speed and frequency. Maybe you could run the GPU memory stress test to see if memory usage will go up at all, and up to what point. This will also tell you if there is a memory or memory controller/bus problem.

I really don't think it's the PSU's fault, because typically when a PSU is bad you get overall system instability and/or unexpected/random shutdowns/restarts.

 

[Lazermonkey]

What does the initial GPU-Z screen say? Perhaps your PCI-E link speed has dropped somehow..? The GPU is not being fed information properly, this is why your VRAM usage is low, and why the GPU is so cool - it's vastly underutilized.

Perhaps you could flash a new motherboard BIOS, and check it's settings to verify PEG link speed and frequency. Maybe you could run the GPU memory stress test to see if memory usage will go up at all, and up to what point. This will also tell you if there is a memory or memory controller/bus problem.

I really don't think it's the PSU's fault, because typically when a PSU is bad you get overall system instability and/or unexpected/random shutdowns/restarts.

What would make my BUS speed drop like that? it will fluctuate between those two numbers in the pics above.

 

[McSteel]

It's probably a power saving "feature"... Ugh. Try setting the power profile to High Performance, or manually disable PCI-E link state power saving in the advanced power settings. The trouble with this is that sometimes the motherboard/GPU don't re-initialize the link to run it at full speed, and you get performance issues. Especially so after long idle periods.

If this doesn't help, you could try putting the card in a different PCI-E slot. Running the latest BIOS is good practice, in my opinion.

 

[Lazermonkey]

It's probably a power saving "feature"... Ugh. Try setting the power profile to High Performance, or manually disable PCI-E link state power saving in the advanced power settings. The trouble with this is that sometimes the motherboard/GPU don't re-initialize the link to run it at full speed, and you get performance issues. Especially so after long idle periods.

If this doesn't help, you could try putting the card in a different PCI-E slot. Running the latest BIOS is good practice, in my opinion.

My motherboard is maybe a month old. I installed the latest BIOS.

I also tried the other PCI-E slots and nothing helped.

Could it be my motherboard is bad?

All of the power saving functions are disabled.

this issue came out of no where.

 

[McSteel]

Well, try running this and see how it goes.

If the test can't allocate the VRAM properly, it might even be a faulty VGA. Also, you could have a memory address space conflict somewhere, but that's a stretch, and harder to actually pinpoint... We'll know more after the test.

 

[Lazermonkey]

Well, try running this and see how it goes.

If the test can't allocate the VRAM properly, it might even be a faulty VGA. Also, you could have a memory address space conflict somewhere, but that's a stretch, and harder to actually pinpoint... We'll know more after the test.

Okay back to square one and then some.

I plugged in a 32GB USB thumb drive and it fried the thumb drive.

I tried an 8GB drive, same thing.

Something is majorly wrong here.

 

[McSteel]

What do you mean "fried"? Confirmed not working in another PC after being plugged into the problematic one, or was there any actual damage involved (no matter how severe)?

Sounds bad for the PSU, which I was sure was in the clear... Though most motherboards power at least some of the USB ports off of 5VSB rail, which doesn't actively power any other component when the computer is on (the regular 5V rail takes over). Did you use the case's front USB ports? Perhaps they're connected/wired wrong? Check that ASAP.

 

[Lazermonkey]

What do you mean "fried"? Confirmed not working in another PC after being plugged into the problematic one, or was there any actual damage involved (no matter how severe)?

Sounds bad for the PSU, which I was sure was in the clear... Though most motherboards power at least some of the USB ports off of 5VSB rail, which doesn't actively power any other component when the computer is on (the regular 5V rail takes over). Did you use the case's front USB ports? Perhaps they're connected/wired wrong? Check that ASAP.

I have discovered the front USB ports are faulty.

I also think the situation with the ports screwed up the board.

Going to Microcenter today to return the board and the case.

Really am upset it killed two USB thumb drives.

 

[Lazermonkey]

Brand new case Antec GX 700. So far so good with the USB front ports.

Brand new Motherboard. ASrock 970 Extreme 4.

I still have the issue with my graphics card not using more than 84-86mb of memory and switching between PCI-E 1.1 and 2.0

I guess the graphics card got damaged when the USB port shorted?

Graphics card is going to MSI this week for RMA.

This sucks.