Driver for GMA X3000 (HW T&L)

[directcpp]

Hi.
please.
When Intel will be released drivers for GMA X3000 (HW T&L and vertex shader 2.0 or 3.0)?
in February?

[noxxle]

Why arent the drivers available on the mobile chipset site yet?

[tcdelaney]

Two things would be my guess.

1. Lots of pages to update, including the knowledgebase for each chipset, etc.

2. By putting them on individual motherboard sites first, they have a slower uptake of the drivers, so if there is a showstopper it affects fewer people before they find out about it.

[jaypizzle]

The desktop drivers are usable for the mobile X3100 too though?

[Aaron_B_Intel]

Yes. Intel has "unified" drivers which work on everything from 945G onward.

The only reason these drivers would not work is if your specific laptop vendor has made them not work (with the intention of providing their own, modified driver).

[noxxle]

Arch, are the devs aware that the 15.4.4 vista driver gives much better performance for many games than the 15.7? In Galciv2, Homm: tribes of the east, and simcity socities I get over double the FPS with 15.4.4. In most other games that run under both the 15.4.4 and 15.7 driver, the 15.4.4 is often slightly faster too.

[Aaron_B_Intel]

Have you tried forcing software TnL? By default the drivers are currently using HW TnL, and in some cases the way the software is written makes itmore efficient to run in SW with a powerful CPU than use the chipset HW.

To enable SW TnL for a specific game, open the registry, do a search for

_3DMark06.exe

In that same area, create a new DWORD entry: "_[name of executable for the game]"

Set it to 1 for SW TnL. Swap it back to 0 for HW TnL.

Note that there are like a billion subdirectories (okay more like 12) which contain _3DMark06.exe and you'll have to create the same new entry in each one, just to be sure.

[noxxle]

that worked, thank you. But I only had to add the new dward in the first reg entry that came up. Do you think a 'profile' manager could be included in the next driver release so users can control which games use HW shaders?

*edit* Also, i noticed that the games that had graphical texture problems in 15.4.4 still have them when forcing software mode under 15.7. Will the devs be doing anywork on the software mode compatibility in the future?

[Aaron_B_Intel]

Unfortunately, I just don't know. I'm in manufacturing. :) I just happen to watch these boards and try to help out where I can.

[vak]

well, i can't find the realeased drivers neither for mobile nor for desktop chipset. Anyone can post the links? thanks

[Aaron_B_Intel]

XP:

http://downloadcenter.intel.com/filter_results.aspx?strTypes=all&ProductID=2412&OSFullName=Windows*+XP+Professional〈=eng&strOSs=44&submit=Go%21

Vista32:

http://downloadcenter.intel.com/filter_results.aspx?strTypes=all&ProductID=2412&OSFullName=Windows+Vista*+32〈=eng&strOSs=164&submit=Go%21

Vista64:

http://downloadcenter.intel.com/filter_results.aspx?strTypes=all&ProductID=2412&OSFullName=Windows+Vista*+64〈=eng&strOSs=150&submit=Go%21

[noxxle]

Why is it taking so long for these drivers to appear on the mobile chipset downloads section? Could they be fixing bugs for a 15.7.1/14.32.1 release instead?

[Aaron_B_Intel]

This happens pretty much all the time, or used to. There's nothing to it really other than a proliferation delay. The folks who maintain the website are not the same ones who released the drivers and perhaps the message only went out (so far) to the motherboard guys.

[davidc1]

I wouldn't doubt there is a problem with the 15.7 and 14.32 drivers as Noxxle suggests. You could be right Archibael, but many more people than I thought are having "default screen resolution" problems with the 14.32 drivers. I am one of the lucky people that doesn't have the problem. Soon after the "game-changing"(pardon the pun) 14.31 production drivers, 14.31.1 drivers were released which fixes a major bug on the 14.31 driver, not being able to switch between hardware/software mode properly.

Intel is nowhere near the "stable" level of the drivers yet. They still need a lot more effort.

[Aaron_B_Intel]

Put it this way: 14.32.2 does not fix the "Default resolution problem". They're not holding up 14.32 to introduce that one. And my source has no 15.7.x yet (if there's one on the roadmap, it's not yet released internally).

If there were a problem with 14.32 and 15.7 which made these guys think they needed to stop distributing them, they'd yank them from the motherboard sites as well.

[noxxle]

archibael it just seems absolutely ridiculus that some of these 'simple' bugs have not been squashed. Exactly how 'underfunded' is the team working on these drivers? Would it be possible to communicate with them? Could someone direct them to these forums or something? Perhaps there are good reasons why some of these issues have not been resolved and they can explain it to us.

[Aaron_B_Intel]

You've got no argument from me; something as obvious as the "default to 800x600" bug should have been found-- if not by Intel's validation labs, certainly by the motherboard and PC vendors who had the driver for XY weeks to evaluate before it went Production Version.

I've got no answer for the rest of your queries. I don't know if the driver team follows this Forum (closely or otherwise), but I suspect they don't want to deal with the venom. In a way I can sympathize: it's easy for me to sit here and chat back and forth with you guys: it's just my employerand not my particular work being criticized. It stings for me, but not as much.

[davidc1]

archibael:

Put it this way: 14.32.2 does not fix the "Default resolution problem". They're not holding up 14.32 to introduce that one. And my source has no 15.7.x yet (if there's one on the roadmap, it's not yet released internally).

If there were a problem with 14.32 and 15.7 which made these guys think they needed to stop distributing them, they'd yank them from the motherboard sites as well.

I don't have that "Default Resolution" problem with XP Service Pack 2 and 14.32 drivers. I think the reason they didn't figure out the problem is because it might be just an intermittent problem-it doesn't appear with all computers.

[tcdelaney]

FWIW, I've had confirmation that the issueI reported (scaling at native resolution) is being investigated, and will be part of a new driver at some stage.

I would think that the 800x600 problem would be considered a higher priority since there's a simple workaround for my issue, but it could well be that they just haven't been able to replicate thethatissue yet.

[prashu162]

14.32 drivers have disappeared from the downloads section.Where did the drivers go?????

[Aaron_B_Intel]

Duplicate Post

[Aaron_B_Intel]

archibael:

Put it this way: 14.32.2 does not fix the "Default resolution problem". They're not holding up 14.32 to introduce that one. And my source has no 15.7.x yet (if there's one on the roadmap, it's not yet released internally).

If there were a problem with 14.32 and 15.7 which made these guys think they needed to stop distributing them, they'd yank them from the motherboard sites as well.

Looks like we were both right, DavidC1