Geforce 600 Series Gen3 Support On X79

[cx-ray]

GeForce GTX 680 and GTX 670 GPUs support PCI Express 3.0. They operate properly within the SIG PCI Express Specification and have been validated on multiple PCI Express 3.0 platforms. Some motherboard manufacturers have released an updated SBIOS to enable the Intel X79/SNB-E PCI Express 2.0 platform to run at up to 8GT/s bus speeds. We have tested GeForce GTX 680 and GTX 670 GPUs across a number of X79/SNB-E platforms at 8GT/s bus speeds, but have seen significant variation in signal timing across different motherboards and CPUs. Therefore we’ve decided to only support and guarantee PCI Express 2.0 bus speeds on X79/SNB-E with our standard release drivers. Native PCI Express 3.0 platforms (like Ivy Bridge) will run at 8GT/s bus speeds with our standard release drivers.

Though we cannot officially support or guarantee 8GT/s speeds for X79/SNB-E, you can manually enable faster 8GT/s bus speeds (at your own risk) on GTX 680 or GTX 670 GPUs using the following executable: force-enable-gen3.exe

If you would like to revert back to 5GT/s bus speeds you can do so by running the following command from the DOS command line: “force-enable-gen3.exe –revert”

You may download "force-enable-gen3.exe" from the File Attachments section below. You will need to reboot your system for the changes to take effect.

http://nvidia.custhelp.com/app/answe...QvaDEzbE45X2s=

[paulbagz]

So no decent 6-core + 6XX support atm?

-PB

[cx-ray]

I don't have the knowledge to see what the Nvidia executable exactly does. However, from what I can tell it only adds a RMPcieLinkSpeed key to the registry entry of an installed GPU. In that respect it's the same as the PCIe 3.0 registry hack that has been posted on several forums.

I think they do support 6 cores. In BF3 I'm almost constantly maxing out (97-98%) 3 GTX 680 cards.

[SKYMTL]

I know the incompatibility of some AMD cards I have tested has been traced back to an X79 PCI-E 3.0 timing issue. It only pops up once in a while with my X79 WS but it has happened to me twice now with two different AiB custom cards. It makes me wonder why AMD is enabling PCI-E 3.0 speeds on their products.....

[safan80]

I know the incompatibility of some AMD cards I have tested has been traced back to an X79 PCI-E 3.0 timing issue. It only pops up once in a while with my X79 WS but it has happened to me twice now with two different AiB custom cards. It makes me wonder why AMD is enabling PCI-E 3.0 speeds on their products.....

I have the trouble on my asus Z9PE-D8 WS which is c602 based and I've been bugging Asus to fix this. Bios update 0602 that just came out yesterday 6/21 still doesn't resolve this on with sapphire 7750 or sapphire 7970 both pci-e 3.0 gpus. The problem was fixed a while back on my sabertooth x79 and no longer exists I can use the same memory/gpus and xeon cpus on my x79 but not my c602

So no decent 6-core + 6XX support atm?

-PB

You should change that to no decent 28nm pci-e 3.0 gpu support

[SKYMTL]

I have the trouble on my asus Z9PE-D8 WS which is c602 based and I've been bugging Asus to fix this. Bios update 0602 that just came out yesterday 6/21 still doesn't resolve this on with sapphire 7750 or sapphire 7970 both pci-e 3.0 gpus. The problem was fixed a while back on my sabertooth x79 and no longer exists I can use the same memory/gpus and xeon cpus on my x79 but not my c602

Personally, I think AMD should go the NVIDIA route: disable PCI-E 3.0 on X79 boards. But then again, it is a marketing point so who knows if that will happen.

[boxleitnerb]

Yup, better safe than sorry. It's not Nvidias fault for sure. Intels chipsets have been quite a disappointment lately.

[Iconyu]

Yup, better safe than sorry. It's not Nvidias fault for sure. Intels chipsets have been quite a disappointment lately.

The PCIE3.0 controller is on the CPU, not the chipset, and Intel didn't have anything to test 3.0 with otherwise they would have had it running on LGA2011 correctly.

[bill_d]

Personally, I think AMD should go the NVIDIA route: disable PCI-E 3.0 on X79 boards. But then again, it is a marketing point so who knows if that will happen.

it has worked fine for me on my 7970's and a asus R4E, but because it doesn't work on some MB's or some cards for some poeple they should turn it off

[SKYMTL]

The PCIE3.0 controller is on the CPU, not the chipset, and Intel didn't have anything to test 3.0 with otherwise they would have had it running on LGA2011 correctly.

They had several LSI PCI-E 3.0 cards for compatibility testing among many other items. Don't think for a second that Intel didn't have the ability to properly test PCI-E 3.0 during the X79 development cycle.

[Iconyu]

They had several LSI PCI-E 3.0 cards for compatibility testing among many other items. Don't think for a second that Intel didn't have the ability to properly test PCI-E 3.0 during the X79 development cycle.

These 680 cards are the only cards I've heard about with issues, I've not heard about any problems with raid cards, or HD7**0 cards. When these boards came out and Intel wouldn't certify their lga2011 chips for pcie3 all the reviews weren't claiming it was broken. Also you didn't seem concerned about this possible issue when you tested the 7970 here when it comes down to it, Intel couldn't get pci-e 3.0 certified and a few raid cards by one manufacturer probably wouldn't be enough for this certification.

As one of the few people with the resources to back a claim like that, I'd like to see you do so, and I'll read your multi page article when you're done.

[paulbagz]

People get so hostile so quickly on these forums.

I think what SKYMTL said earlier is very true, no way Intel were just in the dark about PCI-E 3.0.

-PB

[JohnZS]

Intel also had a similar issue with PCI-E 2.0 on the X38
All nVidia forceware drivers newer than 190 disabled PCI-E 2.0 on X38
John

[SKYMTL]

These 680 cards are the only cards I've heard about with issues, I've not heard about any problems with raid cards, or HD7**0 cards. When these boards came out and Intel wouldn't certify their lga2011 chips for pcie3 all the reviews weren't claiming it was broken. Also you didn't seem concerned about this possible issue when you tested the 7970 here when it comes down to it, Intel couldn't get pci-e 3.0 certified and a few raid cards by one manufacturer probably wouldn't be enough for this certification.

As one of the few people with the resources to back a claim like that, I'd like to see you do so, and I'll read your multi page article when you're done.

No, I wasn't concerned because none of this was out in the open, nor had additional cards been tested, nor do I typically comment about rumors when posting an article. However, as I have tested additional AMD cards, it is becoming painfully obvious that there ARE issues. Significant issues as well. Some members that posted in this very thread have experienced problems with their AMD cards and the Game.AMD forums are littered with people having compatibility hiccups on the X79 platform.

It is almost impossible to replicate the issue on a constant basis. Testing for it and nailing it down requires much more electrical and engineering knowledge than I current have.

One other thing which will be added to all reviews going forward: ALL cards except the dual GPU ones are tested at PCI-E 2.0 speeds.

[JohnZS]

It is looking like this could be a potential errata/defect with the i7-38xx and i7-39xx line of CPUs

Have a look at the following information on intel's site and see what you think?

Intel Xeon E5 Specifications

Now click "Expansion Options"

PCI Express Revision
3.0
# of PCI Express Ports
40

Intel Core i7-3960K Extreme Edition Specifications

Now click "Expansion Options".

PCI Express Revision
2.0
# of PCI Express Ports
10

BOTH CPU's work on the X79 motherboard, however the Xeon E5 clearly states that is has PCI Express 3.0, yet the Core i7 only supports PCI-Express 2.0

This could potentially be microcode releated, so a BIOS update could resolve the issue, however if this BIOS update does not resolve the PCI-E 3.0 issue, then I guess we will need to await a new Stepping C3? D1? of the CPU to hopefully included a fixed PCI-E Controller.

John

[Nizzen]

Callsign_Vega her on XS proved that pci-e 3.0 works with the "regfix" on x79. The gain was pretty good with 680 in quad sli and surround gaming. Bf3 was tested.

[SKYMTL]

No one is saying that is doesn't work. There are huge stability issues though.

[Ice009]

No one is saying that is doesn't work. There are huge stability issues though.

Have you done much testing with it? What are the stability issues? Just random crashes or is in game performance hindered?

Kinda annoyed at this because I bought the Rampage 4 Formula so I can use 2-3 vid cards in Surround. Do you think PCI-e 2.0 would be enough for 2 cards in Surround 5760 x 1200? Not sure if it will be enough for 3 or 4 cards though.

[STEvil]

For 1-2 cards you're fine. Bandwidth starts to become an issue at 3 and 4 cards, mostly only 4.

[Natalia]

For 1-2 cards you're fine. Bandwidth starts to become an issue at 3 and 4 cards, mostly only 4.

I have an ASRock X79 Extreme6/GB and three EVGA 670 FTW's arriving tomorrow, so I can test this little "app" and see what happens. Not sure how many people have tried it on the type of hardware I have coming...

[Chrono Detector]

Wow, having PCI-E 3.0 disabled at higher resolutions does make a difference in FPS, Intel or NVIDIA or whoever is responsible for this mess really need to fix this issue ASAP, whats the point of having PCI-E 3.0 on the motherboard and its been disabled is a waste.

[BoredByLife]

I know the incompatibility of some AMD cards I have tested has been traced back to an X79 PCI-E 3.0 timing issue. It only pops up once in a while with my X79 WS but it has happened to me twice now with two different AiB custom cards. It makes me wonder why AMD is enabling PCI-E 3.0 speeds on their products.....

Shouldn't it be, I don't understand why Intel bricks PCI-E 3.0 but says it supports it? Nvidia or AMD or hardly to blame for this.