Re: Inspiron 1525 w/ GM965 does NOT like new Intel drivers -- t - Page 4

sargeras9 · ‎10-03-2008

UPDATE (12/22/08):
- 7.15.10.1537 is last working version, 7.15.10.1545 and later have the problem
- Problem only affects notebook monitor, not external displays (via VGA connection)
- adjusting brightness and gamma to certain levels fixes problem, but results in extremely bright screen (so, doesn't really fix problem)
- pictures taken from camera attached to this post, where newx.jpg and oldx.jpg correspond to the screen with the newer driver versions and 7.15.10.1537, respectively.
/UPDATE

Hello, first some specs:

Dell Inspiron 1525
Intel Core 2 Duo T5750 2.0GHz
3GB RAM
Vista Home Premium 32-bit
Mobile Intel 965 Express Chipset Family

Alright, now onto the interesting stuff!

So, I bought this laptop, and the display driver version that comes from Dell is 7.14.10.1409 - from Jan 2008. Well, that couldn't do, as Intel has been very good with updating this chipset, giving it vast performance enhancements and better support for games. So, I went to Intel's site to get the latest driver - at the time, it was 7.15.10.1537. So I installed it - it gave the whole "this isn't supported, get the driver from the manufacturer" routine, but after uninstalling the old driver and trying again, it worked like a charm. Without any surprises, my games had a HUGE performance boost (despite the fact that the Vista experience rating dropped from 3.4 to 3.1....pff), so I was a happy man.

Now things get ugly.

So, today I decided it's been a while so I should check if there have been more recent drivers - turns out Intel pushed out 2 newer versions, first 7.15.10.1545 in August, and the more recent 7.15.10.1554 a couple weeks ago. Having read the changelog, I found some nice support features, especially for OpenGL 2.0, so I was happy about getting some new drivers. So I download and install the more recent one with the .exe first. I immediately knew something was wrong - the colors on screen looked like they were melting and were supersaturated, like someone had just spent an hour eating skittles then threw up on the screen. Naturally, it was hard to see or do much in this condition, but I managed to go through some display options, thinking maybe some tweaking was necessary.

I tried everything.
Resolutions, color complexity, messing around with contrast, saturation, etc, you name it. Just for kicks, I took a screenshot and sent it to one of my friends, and he said it looked fine, nothing was wrong. Clearly, this driver was screwed up pretty badly... Well, I installed 7.15.10.1537 again, and everything returned to normal - no problems. Now I thought, well, if the most recent version is messed up, I'll try the other one - 7.15.10.1545. Installed it, same exact thing happened.

I've been going back and forth, trying all methods of install (.zip extraction, executable, update through device manager, etc) and same results. I've nearly spent all day going through Intel's site and just seeing what I can find on Google, but to no avail (Intel's support hasn't been any help, either).

I come here to you now, exhausted, asking for any suggestions anyone may have to fix the problem - it could very well be that the last two drivers are bad, but I seriously doubt it - I've seen other threads where people are using it just fine.

I've ruled basically everything I can think of - including all "dumb" stuff like corrupt files, right version for OS, etc. Will these two simply not work on my machine? I hope the trend doesn't continue with new releases... maybe I can just skip these two and wait for next month's?

tl;dr
Dell came with 7.14.10.1409
I installed 7.15.10.1537 from Intel, works amazing
Tried 7.15.10.1545 and 7.15.10.1554, colors completely screwed up on screen

Thanks in advance, guys. I appreciate anything you can think up on this.

RytoX · ‎03-30-2009

Quoting - sargeras9

I downloaded them but haven't had a chance to install yet, I'll update this when I do. I wouldn't hold your breath, as the last 6 driver releases all had the problem. No, Intel and Dell have not acknowledged the issue, in fact Intel dismissed it by saying the problem could not be reproduced (obviously they didn't use this exact model of laptop).

I confirmed in a previous post that using an external monitor works fine with the newer drivers; it's only a problem with the laptop screen.

Grabbed the zip file for 15.12.4.1666 (7.15.10.1666) from http://downloadcenter.intel.com/filter_results.aspx?strTypes=all&ProductID=2800&OSFullName=Windows+Vista*+Home+Premium%2C+32-bit+version〈=eng&strOSs=153&submit=Go! and installed it last night. I had the laptop hooked up to and running an extended monitor just in case something went horribly wrong. However, it seems to have installed without causing any abnormalities, even after removing the second monitor from the system and rebooting.

I installed by unzipping the driver package and running the included Setup file. It installed, and seems to be working just fine. Sadly, I haven't noticed any improvements in performance (in fact, the graphics scores in the Experience Index dropped from 3.5 to 3.4), but at least the driver seems to work on my Inspiron 1525. Has anyone else gotten this new driver to work on an Inspiron 1525?

My system:
Dell Inspiron 1525
Intel Core2 Duo T5800 @ 2.00GHz
4.00 GB RAM
Vista Home Premium 32-bit
Mobile Intel 965 Express Chipset Family (Mobile Intel GMA X3100)

RytoX · ‎04-20-2009

Quoting - RytoX

Grabbed the zip file for 15.12.4.1666 (7.15.10.1666) from http://downloadcenter.intel.com/filter_results.aspx?strTypes=all&ProductID=2800&OSFullName=Windows+Vista*+Home+Premium%2C+32-bit+version〈=eng&strOSs=153&submit=Go! and installed it last night. I had the laptop hooked up to and running an extended monitor just in case something went horribly wrong. However, it seems to have installed without causing any abnormalities, even after removing the second monitor from the system and rebooting.

I installed by unzipping the driver package and running the included Setup file. It installed, and seems to be working just fine. Sadly, I haven't noticed any improvements in performance (in fact, the graphics scores in the Experience Index dropped from 3.5 to 3.4), but at least the driver seems to work on my Inspiron 1525. Has anyone else gotten this new driver to work on an Inspiron 1525?

My system:
Dell Inspiron 1525
Intel Core2 Duo T5800 @ 2.00GHz
4.00 GB RAM
Vista Home Premium 32-bit
Mobile Intel 965 Express Chipset Family (Mobile Intel GMA X3100)

Bumping the thread. Anyone?

darkdemon135 · ‎04-21-2009

Quoting - RytoX
Bumping the thread. Anyone?

I don't use Vista, and i have no intention to. But i have the same issue on my inspiron 1525. And there is still no solution for winXP. Well, anyway there is no way to play games on X3100 obviously, and i don't believe that some new drivers will help this GMA to run 2005+ games.

sargeras9 · ‎04-21-2009

Sorry for the long wait, everyone.

I tested out 1666, and as expected, the problem has not been fixed. If it worked for you, look at the difference between out specs and read what I said before - it seems the problem only occurs on the Inspiron 1525 with my specs - the ones with your's don't have it. I guess they changed something with our batch that is not present in the others.

This is the seventh driver package release since 1537, and the issue still has not been addressed. As such, I don't think it ever will be unless specifically stated in the release notes for a driver, so I'm not going to keep testing each release unless there is reason to believe the problem has been fixed.

As for the post above, although I don't game on this laptop anymore (I built myself a nice desktop) it could handle Half-Life 2 and Team Fortress 2 decently, so don't count us gamers out. The 1537 driver package was MUCH better than the 1409 (the version Dell recommends) which couldn't play anything at all. This proves the drivers can make a big difference, so don't dismiss this issue as if it doesn't matter.

RytoX · ‎04-21-2009

Quoting - sargeras9

Sorry for the long wait, everyone.

I tested out 1666, and as expected, the problem has not been fixed. If it worked for you, look at the difference between out specs and read what I said before - it seems the problem only occurs on the Inspiron 1525 with my specs - the ones with your's don't have it. I guess they changed something with our batch that is not present in the others.

This is the seventh driver package release since 1537, and the issue still has not been addressed. As such, I don't think it ever will be unless specifically stated in the release notes for a driver, so I'm not going to keep testing each release unless there is reason to believe the problem has been fixed.

As for the post above, although I don't game on this laptop anymore (I built myself a nice desktop) it could handle Half-Life 2 and Team Fortress 2 decently, so don't count us gamers out. The 1537 driver package was MUCH better than the 1409 (the version Dell recommends) which couldn't play anything at all. This proves the drivers can make a big difference, so don't dismiss this issue as if it doesn't matter.

I'm not an engineer, but I don't think the difference between our systems is significant enough. My processor is just barely a series above yours (T5750 vs T5800) and I have one more gigabyte of RAM (3GB vs 4GB). We both have the same model laptop, the same BIOS, the same video card, and (very likely) the same motherboard (though we'd have to check to confirm that). Furthermore, you didn't say that it is specifically only your system, though you did say that it was a possibility (meaning it was not confirmed).

My laptop was purchased around December 12, 2008. Could there be another hardware difference between our laptops that we are unaware of? Could there be a hardware revision difference between our systems? If that's the case, we should start comparing those to see if we can't find where the problem lies. I have seen conflicts between audio cards, motherboards, and video cards crash PCs repeatedly at random. In the meantime, I will check with some friends who also have Inspiron 1525 laptops to see if any of them have the exact same specs as your laptop.

Ah, to answer a previous question from 7oby: The igfxcfg.exe > Information box says that I am running VBIOS 1566.0 with driver version 7.15.10.1666.

7oby · ‎04-21-2009

Quoting - RytoX

My processor is just barely a series above yours (T5750 vs T5800)

The only difference is the FSB of 667 instead of 800 that RytoX and me (on M1330) are using. Sounds not very likely, but maybe one can check if there's one user having a FSB800 CPU and the same issue on the inspiron.

and I have one more gigabyte of RAM (3GB vs 4GB).

Similar for me. I used 2GB and 4GB. The difference between 3GB and 4GB ist actually "flex mode". Flex mode is the ability of the GM965 chipset to drive a 2GB + 1GB configuration with 2x 1GB running in dual channel mode and the remaining one in single channel mode.

If you check the specification update of the GM965 chipset you'll find the following errata:
"Visual Corruption with Integrated Graphics and Intel Flex Memory Technology Enabled using Microsoft Windows Vista Operating System"
http://www.intel.com/Assets/PDF/specupdate/316274.pdf

However according to the same specification update this should be fixed in driver revision v15.4.3. I can hardly imagine this bug got reintroduced. Furthermore I'm convinced this bug is the same as the one described in this blog (warning real low level complicated stuff):
http://keithp.com/blogs/intel_gfxcon_2008/
more information in the linux source code comments regarding this issue (swizzle):
http://lxr.linux.no/linux+v2.6.29/drivers/gpu/drm/i915/i915_gem_tiling.c

And I know how this bug looks like. It's different. It's really corruption - not wrong colors. Anyway: If someone feels bored he might try to pull out the 1GB and force single channel configuration this way.

Could there be a hardware revision difference between our systems?

According to the specification update above there's just one stepping of the GM965. It's C0. At least this I would rule out.

RytoX · ‎04-22-2009

Quoting - 7oby

Quoting - RytoX

and I have one more gigabyte of RAM (3GB vs 4GB).

Similar for me. I used 2GB and 4GB. The difference between 3GB and 4GB ist actually "flex mode". Flex mode is the ability of the GM965 chipset to drive a 2GB + 1GB configuration with 2x 1GB running in dual channel mode and the remaining one in single channel mode.

If you check the specification update of the GM965 chipset you'll find the following errata: "Visual Corruption with Integrated Graphics and Intel Flex Memory Technology Enabled using Microsoft Windows Vista Operating System"

http://www.intel.com/Assets/PDF/specupdate/316274.pdf

However according to the same specification update this should be fixed in driver revision v15.4.3. I can hardly imagine this bug got reintroduced. Furthermore I'm convinced this bug is the same as the one described in this blog (warning real low level complicated stuff):
http://keithp.com/blogs/intel_gfxcon_2008/
more information in the linux source code comments regarding this issue (swizzle):
http://lxr.linux.no/linux+v2.6.29/drivers/gpu/drm/i915/i915_gem_tiling.c

And I know how this bug looks like. It's different. It's really corruption - not wrong colors. Anyway: If someone feels bored he might try to pull out the 1GB and force single channel configuration this way.

Quoting - RytoX

Could there be a hardware revision difference between our systems?

According to the specification update above there's just one stepping of the GM965. It's C0. At least this I would rule out.

So, are you saying that the bug that sargeras is reporting is the same one as on Keith's blog, but different from the "Visual Corruption" in the spec? Are you suggesting that sargeras pulls out his 1GB stick of RAM to force Flex mode to turn off? Alternately, could it be tested by replacing one of my 2GB sticks with a 1GB stick to turn Flex mode on? Lastly, my Information shows "Device Revision: 0C".

7oby · ‎04-22-2009

Quoting - RytoX

So, are you saying that the bug that sargeras is reporting is the same one as on Keith's blog, but different from the "Visual Corruption" in the spec?

The bug in the spec "Visual Corruption" that's not further described is the same problem Keith describes in his blog. They had the problem with the Linux drivers as well. There is a chance - but an unlikely one that sargeras problem vanishes if flex mode is not used.

Are you suggesting that sargeras pulls out his 1GB stick of RAM to force Flex mode to turn off?

yes

Alternately, could it be tested by replacing one of my 2GB sticks with a 1GB stick to turn Flex mode on?

yes

Lastly, my Information shows "Device Revision: 0C".

mine too and I expect sargeras as well. The reporting in the GUI is taken from

PCIVEN_8086&DEV_2A02SUBSYS_02091028&REV_0C

SUBSYS might be different in different brand computers. Maybe that revision is related to the revision printed in the spec. One could see that be further looking into the spec. But it seems not worth it: There's just one revision of this chip regardless which revision number it carries.

jaaypee · ‎09-25-2009

Hello! First off, thank you for the very insightful thread. Unfortunately, I come with no solutions, but rather a request.

Do you happen to still have the 7.15.10.1537 driver download? Intel has removed the download, and I have searched the web for two hours now, but to no avail. If you are able and willing, please send the download to my e-mail: jp.fLip2@gmail.com

Thanks!

Hope to hear from you soon.

7oby · ‎09-26-2009

These two links work:

. GFX_Vista32_15.11.0.1537_PV_Intel.exe - Vista 32-Bit Version
. GFX_Vista64_15.11.64.1537_PV_Intel.exe - Vista 64-Bit Version

. enter captcha code and select "Download This Driver For Your Device From Open Drivers"
. on the next screen select the alternative Download Link Backup Download Server

The most recent intel driver version for Vista for the GM965 chipset is 15.12.4.1666 and you say in this Version you still have the color bug. Maybe you want to give Windows 7 a shot. It's latest driver revision for this video chipset is 15.12.75.3.1912.

7oby · ‎09-26-2009

Let me try the links again:

. GFX_Vista32_15.11.0.1537_PV_Intel.exe
. GFX_Vista64_15.11.64.1537_PV_Intel.exe

jaaypee · ‎09-26-2009

I greatly appreciate the prompt reply, but now i've encountered a new problem!

After uninstalling the 7.14.10.1409 driver version through the Device Manager, and running the driver installation for 1537 (32 bit), I getn an error saying: "The driver being installed is not validated for this computer. Please obtain the appropriate driver from the computer's manufacturer. Setup will exit."

I don't understand why it doesn't work. I uninstalled the 7.14.10.1409 driver yesterday and successfully installed the latest driver, 1662 I believe. Of course, I had to revert back to 1409 because of the broken colors. Nonetheless, it successfully installed! Not to mention the OP was able to do it with the exact same computer (Inspiron 1525).

Hoping for another quick reply/fix,

jaaypee

7oby · ‎09-27-2009

. after you execute the above files they will unpack in a seperate folder such as GFX_Vista32_15.11.0.1537_PV_Intel even if installation quits due to the message you posted.
. Open GFX_Vista32_15.11.0.1537_PV_Intelinstallation_readme.txt
. scroll to section MANUAL INSTALLATION INSTRUCTIONS and follow each step precisely, don't shortcut any step and it works.

I posted a graphical guide here. It's a known issue that you are facing.

jaaypee · ‎09-27-2009

It worked! Thanks so much! You saved me hundreds of dollars, as I was considering buying a new laptop since the outdated driver prohibited me from playing this one MMORPG :)

Inspiron 1525 w/ GM965 does NOT like new Intel drivers -- thorough description