GTX 670 Freeze

[Akumasama]

I got another question for all of you.
Does anybody of you have a chipset OTHER than Z77? Like Z68 maybe?
Because when I tested my videocard on 2 different systems, both had Z77 and both presented the freeze issue in PCIE 3.0 16x.

I sent it back to the shop where I bought it, and they say it's working fine in PCIE 3.0 16x on a Z68 chipset motherboard. So I'm starting to wonder if the issue happens only with Z77.
The two motherboards I tested it with are:

Asrock Z77 Extreme 4
Asus Sabertooth Z77

[kveksen]

When I enable pcie 3 in bios my system won't even boot.
GTX 670 windforce in MSI x79 MB...

[crinklecut]

I got another question for all of you.
Does anybody of you have a chipset OTHER than Z77? Like Z68 maybe?
Because when I tested my videocard on 2 different systems, both had Z77 and both presented the freeze issue in PCIE 3.0 16x.

I sent it back to the shop where I bought it, and they say it's working fine in PCIE 3.0 16x on a Z68 chipset motherboard. So I'm starting to wonder if the issue happens only with Z77.
The two motherboards I tested it with are:

Asrock Z77 Extreme 4
Asus Sabertooth Z77

I got a p67a-ud7-b3 pci express gen 2 motherboard, same issue everytime i try and run a game 30 secs to 5 mins into it the screen will go black with no monitor signal, sometimes returning to the game a maximum of once after 5 to 10 secs, only to have the performance of the card severely throttled back like im playing on a 6600gt or something - with the sound always cut out if im using a pci express sound card, if im using the on board realtek sound that still runs but the performance of the gtx 670 is still severely throttled back, then 15 secs later the screen will freeze without fail EVERY time. It's as if the PCI express bus gets over saturated briefly and goes into some power saving mode which disrupts the sound card and throttles back the graphics card, resulting in said crash of sound card and then graphics card 15 secs later...

Also I have to add that I was unaware that my board had 2 half speed x8 pci express slots until after I scrutinised it under the downlights in my room, noticing that 2 of the slots are only half populated with pins, im adding this because I had my card in the lower left hand side said x8 slot when i initially installed it, heaven ran the same framerate in both the x8 and x16 slot which was interesting to note...doesnt exactly point to an over saturation of the pci express bus now does it - seems more like a power throttling issue to me as these cards do dynamically change their frequency in proportion to the workload, did Nvidia screw up this part of the driver release?

[Altair]

My motherboard is a P8Z68 deluxe so my chipset is Z68
At the shop they probably couldn't find the problem because the crashes are random and unpredictable. Sometimes I could run a game for 5 hours before the first crash, I don't think they tested the card for so long.
Other times the crashes happens at desktop within a few minutes after tunring the PC on.

[Akumasama]

For me the crash was always very fast, whenever I turned 3Dmark, Windows Performance Index, GPU-Z's render box or anything using the videocard etc.
I only needed to do that and wait a few mins to see the freeze.

[Altair]

My card never crashed while performing benchmarks. It probably depends on the system. Maybe on Z68 the crashes happens randomly: it tooks me a week to realize there was something more severe then just a bugged game crashing from time to time! For this reason at the shop the tests appeared to be ok.

[Akumasama]

Could be!

Altough I still wonder why only a small part of these cards have this issue. Maybe we should check Serial Numbers and see if they're close to each other? Or would that be useless?

[Altair]

 I don't know how much serial numbers matters. In any case, without support from Gigabyte it is useless to check them.

[darivo]

any official news from gigabyte???

[Akumasama]

any official news from gigabyte???

No I don't think they have officially acknowledged the issue.

I forgot, did you manage to try in PCIE 3.0 8x, or PCIE 2.0? Do you still get freeze issues?
I don't know much the X79 chipset platform. It's a LGA2000 socket? Can it handle Ivy Bridge processors?

[darivo]

any official news from gigabyte???

No I don't think they have officially acknowledged the issue.

I forgot, did you manage to try in PCIE 3.0 8x, or PCIE 2.0? Do you still get freeze issues?
I don't know much the X79 chipset platform. It's a LGA2000 socket? Can it handle Ivy Bridge processors?

with pci16x 2.0 i haven't issues

[Akumasama]

Then appearently you have the same exact issues as me, Altair and the other guys from that Anandtech thread I linked in one of my previous posts.
The issue of the guy who opened this thread was probably different, since appearently he solved it with the F4 bios.

Would be good to have an official reply from Gigabyte. Something like:

• We have acknolwedged this issue and are looking for a solution
• We know of this, a future bios update will solve it
• We would like to study the issue, could some of you please send the card to our offices?

Whatever really.
I'm still worried about what they said to me at the shop where I bought my card, I sent it back in RMA. They said it worked fine for them on PCIE 3.0 16x, but on a Z68 chipset moherboard, while me and Altair have a Z77 and you have an X79.
Either they were lying to me, or this issue doesn't affect all chipsets.
The more details I discover about this problem, the more it seems so insanely strange, bizarre and complicated.
I've been an enthusiast PC user for 2 decades and I also worked both as a PC components reseller and PC HW & SW assitance, and I swear I've never seen anything as strange, situational and "hidden" like this.

[darivo]

Then appearently you have the same exact issues as me, Altair and the other guys from that Anandtech thread I linked in one of my previous posts.
The issue of the guy who opened this thread was probably different, since appearently he solved it with the F4 bios.

Would be good to have an official reply from Gigabyte. Something like:

• We have acknolwedged this issue and are looking for a solution
• We know of this, a future bios update will solve it
• We would like to study the issue, could some of you please send the card to our offices?

Whatever really.
I'm still worried about what they said to me at the shop where I bought my card, I sent it back in RMA. They said it worked fine for them on PCIE 3.0 16x, but on a Z68 chipset moherboard, while me and Altair have a Z77 and you have an X79.
Either they were lying to me, or this issue doesn't affect all chipsets.
The more details I discover about this problem, the more it seems so insanely strange, bizarre and complicated.
I've been an enthusiast PC user for 2 decades and I also worked both as a PC components reseller and PC HW & SW assitance, and I swear I've never seen anything as strange, situational and "hidden" like this.

i have F4 Bios and the problem continue, without a determinate moment of freeze
very strange

[Akumasama]

Try to see if you can replicate the error like me.
If I didn't do anything, the error was random and could happen at any time, even if idling.

If I tried to launch one of the following softwares, the error happened within a few seconds/minutes

1) 3Dmark11
2) GPU-Z's 3D render window
3) Windows' Performance Index benchmark

[darivo]

Try to see if you can replicate the error like me.
If I didn't do anything, the error was random and could happen at any time, even if idling.

If I tried to launch one of the following softwares, the error happened within a few seconds/minutes

1) 3Dmark11
2) GPU-Z's 3D render window
3) Windows' Performance Index benchmark

sorry but no

the error happened to me aleatory, and usually in the desktop (2d mode)