No PCIe 2.0 on recent drivers
  1 / 57    
I sent this problem report to the Driver Feedback couple months ago, the status has been "Researching". I decide to post this to see if anyone also has this problem.

You can check your status using [url="http://www.techpowerup.com/gpuz/"]GPU-Z[/url]. It's under "Bus Interface", mouse-over the box to give detailed info. [url="http://www.sisoftware.net/"]Sandra[/url] and [url="http://www.hwinfo.com/"]HWiNFO32[/url] also give similar results. You can also use [url="http://cuda-z.sourceforge.net/"]CUDA-Z[/url] to test PCIe transfer speed between video card and host system.

FYI before you post:
1. This affects mostly X38 and X48 boards, some Pxx and X58 boards can also be affected.
2. This is NOT cause by power saving on affected boards.
3. To test if you are actually affected by power saving feature, take GPU-Z's reading under load e.g. Furmark.


My spec:
Leadtek 8800GT 512 (PX8800 GT)
Gigabyte EX38-DS4
Intel Core2 Q9450


This is a list of drivers that I've tested ("no" means no 2.0 connection, "yes" means 2.0 connection):

258 and beyond: no 2.0 unless mentioned otherwise
256.xx no
197.45 no
197.13 no
196.21 no
195.62 no
191.07 no
190.62 no
186.18 yes
185.93 yes (also Win 7's built-in driver)
182.08 yes
180.48 no
178.24 yes
178.13 yes

Related links:
[url="http://forums.nvidia.com/index.php?showtopic=175360"]Forum user JohnZS' request about giving option to enable PCIe 2.0 in driver[/url]
[url="http://nvidia-submit.custhelp.com/cgi-bin/nvidia_submit.cfg/php/enduser/std_alp.php"]NVIDIA's driver problem reporting form[/url]
I sent this problem report to the Driver Feedback couple months ago, the status has been "Researching". I decide to post this to see if anyone also has this problem.



You can check your status using GPU-Z. It's under "Bus Interface", mouse-over the box to give detailed info. Sandra and HWiNFO32 also give similar results. You can also use CUDA-Z to test PCIe transfer speed between video card and host system.



FYI before you post:

1. This affects mostly X38 and X48 boards, some Pxx and X58 boards can also be affected.

2. This is NOT cause by power saving on affected boards.

3. To test if you are actually affected by power saving feature, take GPU-Z's reading under load e.g. Furmark.





My spec:

Leadtek 8800GT 512 (PX8800 GT)

Gigabyte EX38-DS4

Intel Core2 Q9450





This is a list of drivers that I've tested ("no" means no 2.0 connection, "yes" means 2.0 connection):



258 and beyond: no 2.0 unless mentioned otherwise

256.xx no

197.45 no

197.13 no

196.21 no

195.62 no

191.07 no

190.62 no

186.18 yes

185.93 yes (also Win 7's built-in driver)

182.08 yes

180.48 no

178.24 yes

178.13 yes



Related links:

Forum user JohnZS' request about giving option to enable PCIe 2.0 in driver

NVIDIA's driver problem reporting form

#1
Posted 12/21/2009 07:46 PM   
Same problem here m8 , I got Foxconn X38A which supports pci-e 2.0 and Zotac GTX260 pci-e 2.0 and GPU-Z 0.3.8 shows that my Graphic card running at pci-e 1.1 WTF I'm asking myself WTF , I have the newest intel drivers ,gforce drivers 195.62, etc.
Same problem here m8 , I got Foxconn X38A which supports pci-e 2.0 and Zotac GTX260 pci-e 2.0 and GPU-Z 0.3.8 shows that my Graphic card running at pci-e 1.1 WTF I'm asking myself WTF , I have the newest intel drivers ,gforce drivers 195.62, etc.

#2
Posted 12/22/2009 03:30 PM   
Actually...
[img]http://gpuz.techpowerup.com/09/12/22/4vq.png[/img]
Seriously... i'm sick of this crap! i just can't take it anymore! Do you guys even know how you OWN PC works? crap!
Every day i'm reading how someone complains about problems that i never face for about 10 years nvidia dedicated user.... i have so much to say to you guys... but i wouldn't tell you anything... just learn to read the forum, there is so many useful guids and stuff that you have to be a retard or some sort of.... to not understand it. On the other side i bet you don't even know how your own system works and what it need.... no more comments!
Actually...

Image

Seriously... i'm sick of this crap! i just can't take it anymore! Do you guys even know how you OWN PC works? crap!

Every day i'm reading how someone complains about problems that i never face for about 10 years nvidia dedicated user.... i have so much to say to you guys... but i wouldn't tell you anything... just learn to read the forum, there is so many useful guids and stuff that you have to be a retard or some sort of.... to not understand it. On the other side i bet you don't even know how your own system works and what it need.... no more comments!

#3
Posted 12/22/2009 06:02 PM   
[quote name='rOO-69' post='968920' date='Dec 22 2009, 03:30 PM']Same problem here m8 , I got Foxconn X38A which supports pci-e 2.0 and Zotac GTX260 pci-e 2.0 and GPU-Z 0.3.8 shows that my Graphic card running at pci-e 1.1 WTF I'm asking myself WTF , I have the newest intel drivers ,gforce drivers 195.62, etc.[/quote]

Try older drivers to see if it's really driver related, because there could be other causes, such as the card is not sitting firmly in the slot, BIOS settings, or other hardware problems, you can also test another PCIe 2.0 slot on the mb. And you need to test if it's not the result of power saving feature, e.g. try running 3D game in window mode while checking GPU-Z.

It takes me a while to know it's the driver. The list is tested with everything else being equal and driver version is the only variable.
[quote name='rOO-69' post='968920' date='Dec 22 2009, 03:30 PM']Same problem here m8 , I got Foxconn X38A which supports pci-e 2.0 and Zotac GTX260 pci-e 2.0 and GPU-Z 0.3.8 shows that my Graphic card running at pci-e 1.1 WTF I'm asking myself WTF , I have the newest intel drivers ,gforce drivers 195.62, etc.



Try older drivers to see if it's really driver related, because there could be other causes, such as the card is not sitting firmly in the slot, BIOS settings, or other hardware problems, you can also test another PCIe 2.0 slot on the mb. And you need to test if it's not the result of power saving feature, e.g. try running 3D game in window mode while checking GPU-Z.



It takes me a while to know it's the driver. The list is tested with everything else being equal and driver version is the only variable.

#4
Posted 12/22/2009 11:37 PM   
Edit your previous post dude...
195.62 no to 195.62 Yes (and maybe the others from "broken" 19x.xx series)
Just be cause someone don't know how to install drivers properly does not mean they are not working :D
Edit your previous post dude...

195.62 no to 195.62 Yes (and maybe the others from "broken" 19x.xx series)

Just be cause someone don't know how to install drivers properly does not mean they are not working :D

#5
Posted 12/22/2009 11:50 PM   
kahnakha, this has nothing to do with them installing the driver wrong. It has happened to quite a few people. Most of those people are running X38 boards from ASUS/Gigabyte and such. IT DOES NOT AFFECT EVERYONE. Some people are running post-186.18 drivers just fine with PCIe 2.0.

Here is an ASUS forum link on the matter...
[url="http://vip.asus.com/forum/view.aspx?id=20091026025645328&board_id=1&model=Maximus+Formula&page=1&SLanguage=en-us"]http://vip.asus.com/forum/view.aspx?id=200...SLanguage=en-us[/url]

I have the ASUS Maximus Formula motherboard mentioned above, and an XFX GTX 260. I also have this very same problem. I noticed one day that GPU-Z and even NVIDIA System Information and other apps were reporting that I was in PCIe 1.1 mode, and not PCIe 2.0 like I should be. So I did quite extensive testing, and came out with much the same results as the person above. I did not test back beyond 182.50 sadly. Every 18x.xx driver worked up until 186.18. Any driver after that, would report PCIe 1.1, all the way up to 195.81.

Supposedly, the people in the above thread have contacted ASUS and Nvidia, and are expecting a fix soon. I hope this is the case.

Pics below...

[attachment=15100:PCIE20_18618.png]
[attachment=15101:PCIE20_19038.png]
kahnakha, this has nothing to do with them installing the driver wrong. It has happened to quite a few people. Most of those people are running X38 boards from ASUS/Gigabyte and such. IT DOES NOT AFFECT EVERYONE. Some people are running post-186.18 drivers just fine with PCIe 2.0.



Here is an ASUS forum link on the matter...

http://vip.asus.com/forum/view.aspx?id=200...SLanguage=en-us



I have the ASUS Maximus Formula motherboard mentioned above, and an XFX GTX 260. I also have this very same problem. I noticed one day that GPU-Z and even NVIDIA System Information and other apps were reporting that I was in PCIe 1.1 mode, and not PCIe 2.0 like I should be. So I did quite extensive testing, and came out with much the same results as the person above. I did not test back beyond 182.50 sadly. Every 18x.xx driver worked up until 186.18. Any driver after that, would report PCIe 1.1, all the way up to 195.81.



Supposedly, the people in the above thread have contacted ASUS and Nvidia, and are expecting a fix soon. I hope this is the case.



Pics below...



[attachment=15100:PCIE20_18618.png]

[attachment=15101:PCIE20_19038.png]

#6
Posted 12/23/2009 11:35 AM   
I checked out 186.18 drivers and GPu-z shows Pci-e 2.0 x16 @ x16 2.0, but newer drivers make it pci-e 1.1 ,kahnakha doesn't know what we r trying to figure out ,
And I'm also on X38 chipset, is that the issue ? and 1 more thing kahnakha this is not only 1 problem I got with gforce drivers , see link[url="http://forums.nvidia.com/index.php?showtopic=153802"]Wolf2MP crashes[/url]
I checked out 186.18 drivers and GPu-z shows Pci-e 2.0 x16 @ x16 2.0, but newer drivers make it pci-e 1.1 ,kahnakha doesn't know what we r trying to figure out ,

And I'm also on X38 chipset, is that the issue ? and 1 more thing kahnakha this is not only 1 problem I got with gforce drivers , see linkWolf2MP crashes

#7
Posted 12/23/2009 12:42 PM   
Yes, it seems mostly tied to X38 and X48 motherboards. Supposedly even Intel X38/X48 boards had this problem at one time, and a BIOS update fixed the issue. Hopefully either ASUS/Gigabyte/others fix this in BIOS, or Nvidia figures out the cause and fixes it, or both.
Yes, it seems mostly tied to X38 and X48 motherboards. Supposedly even Intel X38/X48 boards had this problem at one time, and a BIOS update fixed the issue. Hopefully either ASUS/Gigabyte/others fix this in BIOS, or Nvidia figures out the cause and fixes it, or both.

#8
Posted 12/23/2009 01:19 PM   
Ok i apologize then, i have no idea what are you talking about, maybe because i was never facing it...
However... I think the bios update should do the work just fine... i'm using this mobo [url="http://www.trustedreviews.com/motherboards/review/2008/02/21/MSI-P7N-SLI-Platinum/p1"]http://www.trustedreviews.com/motherboards...SLI-Platinum/p1[/url]
for a year and a half maybe... and till now(and hope for future) i do not have ANY problems nether with drivers, slot configurations or clock speeds.
Ok i apologize then, i have no idea what are you talking about, maybe because i was never facing it...

However... I think the bios update should do the work just fine... i'm using this mobo http://www.trustedreviews.com/motherboards...SLI-Platinum/p1

for a year and a half maybe... and till now(and hope for future) i do not have ANY problems nether with drivers, slot configurations or clock speeds.

#9
Posted 12/23/2009 03:26 PM   
I have the latest BIOS. Some people on that ASUS forum even tried BIOSes dating way back to see if it fixed the issue, it did not.

The reason you are not seeing the issue is probably because you have an Nvidia chipset motherboard, none of those seem to be affected by the "bug." It seems that only Intel X38/X48 chipsets are affected by this. (Unless someone else can provide an example? Haven't seen any so far.)

As for drivers, I have no problems with them either. All the games I have tried work great, GPU acceleration on video works great, etc. The only single "bug" I have seen at all, has been this PCIe 2.0 being reduced to PCIe 1.1 one.
I have the latest BIOS. Some people on that ASUS forum even tried BIOSes dating way back to see if it fixed the issue, it did not.



The reason you are not seeing the issue is probably because you have an Nvidia chipset motherboard, none of those seem to be affected by the "bug." It seems that only Intel X38/X48 chipsets are affected by this. (Unless someone else can provide an example? Haven't seen any so far.)



As for drivers, I have no problems with them either. All the games I have tried work great, GPU acceleration on video works great, etc. The only single "bug" I have seen at all, has been this PCIe 2.0 being reduced to PCIe 1.1 one.

#10
Posted 12/23/2009 04:16 PM   
Hi everyone,
I'm the one that started the post on this issue in ASUS' forum and even contacted their support. If you will read the post, John47 also contacted them and they started to research the problem. This is a driver issue with the X38 chipset and hopefully it will be solved soon.
Thanks for you guys floating this issue up.

Shai /thumbsup.gif' class='bbc_emoticon' alt=':thumbsup:' />
Hi everyone,

I'm the one that started the post on this issue in ASUS' forum and even contacted their support. If you will read the post, John47 also contacted them and they started to research the problem. This is a driver issue with the X38 chipset and hopefully it will be solved soon.

Thanks for you guys floating this issue up.



Shai /thumbsup.gif' class='bbc_emoticon' alt=':thumbsup:' />

[Intel Braiden Creek DZ68BC][Intel Core i7-2600k][Windows 7 Ultimate 64-bit]

[16GB (4x4gb) Kingston DDR3 PC3-12800 HyperX][Dell 2408WFP 24" LCD]

[Indel SSD320 120GB][2x Seagate 3TB Baracuda RAID 0][Gigabyte GeForce GTX 560]

[Creative X-Fi Titanium Fatal1ty Pro][Compucase 6A19B case]

[Corsair HX620 PSU]

#11
Posted 12/23/2009 04:48 PM   
Thanks so much for the info Keen and Shai, it's very revealing.

Here is my story at Leadtek and Gigabyte camp:

Before I reported this problem to Nvidia I did indeed contacted both Leadtek and Gigatybe. To make long story short, they both can reproduce the issue.

Leadtek was using 190.45 and no 2.0. Gigabyte was using 191.07 and 186.18, the old driver has 2.0 but not the new one. It was at that point I realize it has something to do with Nv driver, so kudos to Gigabyte's supports and findings. They both sent me screenshots of their tests.

However here is a little info that ASUS users would like to know: Leadtek originally tested the same card with ASUS P5E, which is also Intel X38 and using 190.45, and it has 2.0.

From Leadtek:
[attachment=15104:from_Lea...EX38_DS4.JPG]
[attachment=15105:from_Leadtek___P5E.JPG]

From Gigabyte:
[attachment=15106:191.07.jpg]
[attachment=15107:186.18.jpg]
Thanks so much for the info Keen and Shai, it's very revealing.



Here is my story at Leadtek and Gigabyte camp:



Before I reported this problem to Nvidia I did indeed contacted both Leadtek and Gigatybe. To make long story short, they both can reproduce the issue.



Leadtek was using 190.45 and no 2.0. Gigabyte was using 191.07 and 186.18, the old driver has 2.0 but not the new one. It was at that point I realize it has something to do with Nv driver, so kudos to Gigabyte's supports and findings. They both sent me screenshots of their tests.



However here is a little info that ASUS users would like to know: Leadtek originally tested the same card with ASUS P5E, which is also Intel X38 and using 190.45, and it has 2.0.



From Leadtek:

[attachment=15104:from_Lea...EX38_DS4.JPG]

[attachment=15105:from_Leadtek___P5E.JPG]



From Gigabyte:

[attachment=15106:191.07.jpg]

[attachment=15107:186.18.jpg]

#12
Posted 12/23/2009 06:22 PM   
PCIe 1.1 or 2.0 is such a non-issue, since all video cards dont even take advantage of the bandwidth that PCIe 2.0 is capable of. Hell, most video cards is barely breaking the PCIe x8 bandwidth.
PCIe 1.1 or 2.0 is such a non-issue, since all video cards dont even take advantage of the bandwidth that PCIe 2.0 is capable of. Hell, most video cards is barely breaking the PCIe x8 bandwidth.

eVGA Z68 SLI | Intel Core i5-3570K @ 4.5 GHz | Corsair Hydro Series H80i
8GB G.Skill Sniper Series DDR3 | EVGA GTX 580 | OCZ ZX 1000W
OCZ Agility3 120 GB SSD + SanDisk Ultra 120GB SSD
Samsung UN55D6000 + Samsung T240
Win8 Pro x64 / WEI - 8.0 - 8.0 - 8.1 - 8.1 - 7.8
3DMark11 - P7545 / Vantage (PPU disabled) - P28859
F@H Team: 142900

#13
Posted 12/23/2009 08:42 PM   
any joy any one ?
any joy any one ?

#14
Posted 12/28/2009 03:50 PM   
Nothing new to report yet. Awaiting new drivers or an updated BIOS to see if either fix it.
Nothing new to report yet. Awaiting new drivers or an updated BIOS to see if either fix it.

#15
Posted 12/29/2009 02:19 PM   
  1 / 57    
Scroll To Top