Fedora Hardware :: Nvidia Developer Driver With CUDA Support?
Sep 29, 2010
I apologize if this is a duplicate post but I was unable to find anything that addressed my question. Currently I've installed the nvidia developer driver (beta) found here:[URL]..
After installing the driver everything works correctly. Rebooting the system results in a hang prior to starting X. Opening a new console and running start X provides me with the following error:
Code:
[ 103.884]
X.Org X Server 1.8.2
Release Date: 2010-07-01
[ 103.885] X Protocol Version 11, Revision 0
[ 103.885] Build Operating System: x86-10 2.6.32-44.el6.x86_64
[Code]..;.
At this point if I reinstall the driver and reissue startx, again the system starts up properly. Also I've noticed that after a reboot, my xorg.conf file is being overwrriten.
I've been trying to get the Nvidia drivers from the nvidia site installed by running-devdriver_4.0_linux_64_270.41.19.run
I have a GeForce 210 PCI Express DDR3 graphics card. I've recently screwed an old version of fedora trying to do this and ended wiping it and installing Fedora 15. I managed to get the standard setup working nice with dual screen support and Tux Kart working smoothly which has never happened before with ATI cards.
The reason i've tried to get an NVIDIA card working is so that i can try to do some development with CUDA. To do this i need to get the proprietary driver working. To get the install program working correctly i've edited the kernel options in grub.conf to stop nouveau driver which seemed to work. The next step i got onto was the need for the kernel source which i've installed. The program can't find this, even with a load of Symbolic links giving-
/usr/src/linux-2.6.40-4 /usr/src/linux /usr/src/2.6.40-4.fc15.x86_64 /usr/src/linux-2.6.40-4.fc15.x86_64 /usr/src/linux-2.6.40-4 all pointing to- /usr/src/Kernels/2.6.40-4.fc15.x86_64
After following the steps of dozens of guides, i now cannot run games that need 3d support and i'm no further in getting the proprietary drivers working. Coding in CUDA is a distant dream after weeks of faff in my spare time to go 2 steps back and one forward...
Does anybody know how to do all this? If anybody could help me i'd be very grateful. I've never had decent graphics support in linux. Everytime i've tried it's ended in a horrible mess!
I have a "stripped down" version of ubuntu 9.10 that was created for me by a developer.It's installed on a Dell Latitude 2100 and I have a flash drive with an image so that I can install it onto multiple Dell Latitude 2100's to use out in the field.The original developer is no longer able to support the software and I have a problem with the gnome power manager (I think that's where the problem is) in that when lid is closed, I get a black screen.I know that the laptop is still running fine behind this black screen because if I plug in an external monitor I can see the screen (only on the external) and use the laptop fine.I know also that the developer stripped out parts of the gnome power manager and possibly acpi as well (if that has anything to do with the lid switch?)I can get the screen back if I go to a terminal and "sudo vbetool dpms on". I do this by using the external monitor but this doesn't solve my issue because out in the field there is no external monitor to use. The lid gets closed often in the field ad the only way to get it back is to hard reboot (hold down the power button till power off and then turn it on again.I've tried removing and re-installing gnome power manager as well as acpi and acpi support but this doesn't help.
I've tried using gconf-editor to set the lid_ac and lid_battery to "blank" and also "nothing" and I still get the same problem.When installing a full copy of 9.10 on these laptops and setting the lid_ac and lid_battery settings to "blank" or "nothing" it works perfectly so it's definitely something to do with what's been stripped out.At this stage, I am just looking for a solution that will allow people out in the field to get the screen back on whenever this happens ... so even if it's a case of them being able to invoke the "sudo vbetool dpms on" command by some manner ... note that part of the stripping has been to disable various key bindings.
It would appear Nvidia are to drop support for the 'nv' driver,recommending using the Vesa driver instead until you install their proprietary driver.Read the article here:
I am looking at purchasing a laptop (HP EliteBook Mobile Workstation), but I am not sure whether to buy one with ATI or NVIDIA graphics. My only real concern for graphics is driver support. Which has better support?
BTW, I dont know if this is important, but I will be installing Fedora linux. 12 or higher.
I've spent about 3 hours yesterday trying to get my GeForce 210 with CUDA DevDriver running on the 3.0.4 kernel. It was running w/o problems on 2.6.39-desktop (with dev-src for 2.6.37). [yeah, I know I should stick to what works - but I don't like it]
After uninstalling the old drivers I tried several ways for an installation of the new ones:
1. ./devdriver_4.0_linux_64_270.41.19.run on the new kernel (even manually navigating to the kernel-src folder)
2. ./devdriver_4.0_linux_64_270.41.19.run on the old kernel, with new dev-src &c. 3. XFree 280.16 driver on the new kernel with the new dev-src &c.
The results:
3. Works now.
2. did not compile telling me the source files were missing (obviously, as they were for 3.0.4 and not 2.6.39)
1. Aborted the installation with the following (excerpt from nvidia-installer.log)
It is important to mention that initially Fedora did see the cards and I was able to set up a dual monitor system. It right was after I enable SLI and PhysX and re-booting into Linux that the problem showed up.
I have seen this issue before in another machine with an ASUS board, but not until today I associated with the SLI setup. My guess is that there has to be something that the driver is enabling in the cards that messes up the interface between the nvidia.ko module and the kernel, but I don't know what may fix it. I need this system for some numerical calculations.
I finishing a install of F14 (i386) on a Laptop with the Nvidia fx 5650 gpu I installed the drivers from the Nvidia site. BUt when trying to set up Desktop effects, I get a Accelerated 3D graphics is not available.
I had effects working in F11 so wondering wjhat I may need to do to get 3d working in F14? I have nouveau blacklisted in modprobe.d and in grub...
it's my understanding, from reading through the forum, that support for older nVidia cards is not included with the official nVidia Kernel Module (kmod-nvidia : RPM Fusion) from what i can tell... it is recommended that users with these cards try the 'nouveau' driver, not used by default in F12 ('nv' is default)... to obtain reasonable functionality.
i'm also under the impression that you need to specify the 'nouveau' driver manually in the xorg.conf... and that an xorg.conf needs to be generated, because F12 does not generate one upon setup. what i'm hoping is to get the nod from experienced users here that my impressions on this are correct.
I have recently been using Fedora, which is awesome, But I have a question, how come I do not have acceleration support with my GPU (Nvidia 9500 m gs), is it not compatible with it? Do I need a special driver?
I'm trying to find out when QME2572 (Qlogic) card became support by the kernel. We have a RHEL 5.1 system that is moving to new hardware, however the kernel at this release doesn't support the new hardware, due to the Qlogic card change. I tired the Redhat KB and Bugzilla. Is there a Kernel change list etc I can search. Never really played around with the kernel too much so I'm just after some pointers for looking up this information. Offically its not supported until Redhat 5.3, I'm trying trying to research kernel info so I can tell the customer they have to upgrade.
My friend just recently gave me an Nvidea Geforce 4 64mb, and it is supposed to be better than my ATI x700 Pro. I was wondering if there was openGL support for this card, and how do i replace it with my ATI card?
I am running a FC14 x64 on a desktop which has a nvidia GeForce310 aka GT218 card which is being supported by the in kernel nouveau driver.
The kernel/syslog reports:
Code:
When Fedora 14 boots, I can hear the nvidia GPU fan spin up to its max rate.
With Windows on this same machine, the same fan starts at the same rate but drops to a lower (and less noisy rate) so we know there is graphics card supports fan control.
Does anyone know if its possible to set the nvidia's fan speed via the nouveau (or other) driver/util?
I have looked at nvclock (0.8b4 being the most recent I could find) but running this reports that the card isn't supported, although it reportedly supports the NV5x backend.
I own an ASUS N61JV-X2 notebook PC that has an Intel HD IGP, nVIDIA GeForce GT 325M GPU, and nVIDIA Optimus technology. I am running OpenSUSE 11.4 64 bit with GNOME. I used to be a former Ubuntu user. I used Ubuntu 10.10 Maverick Meerkat 64 bit. In Ubuntu, it was not recommended that I install the proprietary nVIDIA binary drivers because it would cause me to log into a TTY console upon reboot.
nVIDIA states that they have no plans to support Optimus technology in Linux at any time. Would it be recommended that I install the proprietary nVIDIA binary drivers in OpenSUSE 11.4 64 bit with GNOME? Will I get the same problem as in Ubuntu? How do I do this?
I've installed the Boinc 6.4.5 x86_64 using F10 update package from Fedora repo. The client runs fine using the CPUs, but doesn't fing the nVidia display processors. I've installed nVidia drivers using Fusion repo packages for F10 ( currently 180.25 ) nVidia display settings seems happy and correctly reports driver 180.25 in place as well as 9600 GSO graphics card and LCD monitor.
My objective is to get Boinc using the CUDA capabilities of my GeForce 9600 GSO graphics card and see some of the performance benefits advertised. Although Boinc client is running it will only use the CPU to do work. I have not been able to discover any flags or configuration settings to help Boinc find CUDA library files. I'm stuck at the limits of my knowledge. The following are the startup messages from Boinc. I'm assuming the fact that it can't find the library causes the inability to see the GPU coprocessor.
Being a LAMP snob, it pains me to even think these thoughts, but .I'm running FC-9, and developing websites, mainly using PHP - well the whole LAMP thing. I now find the need to test/code for IE. To that end I'm thinking VMware, VitrualBox, etc. loaded up with that "other" OS.
I'm sure others have gone down this path. I'm guessing the answer is Win-7. My main question is: OK, I do a virgin buy and install and register it on my VM. Six months later I want to change from VM-X to VM-Y and I reinstall - what is MS's feelings on VM's.
I am using a Sparklan WUBR-505N abgn USB dual band stick for the wireless connection on one of my Fedora 12 laptops. I have an Apple Airport Extreme BaseStation that can advertise a 5GHz connection ie it broadcasts two sids... The network manger only see the 'regular' sid and the usb stick only connects in the 2.4GHZ frequencies. Does the rt2870 driver support the 5GHz frequencies? If so how can I force a 5GHz connection?
I am new to linux. I install fedora 10 in text mode in my system. Because some drivers didn't support. I don't no know how to install supporting drivers. My system config. are Intel 945 mother board, 512 DDR2 RAM, 80GB IDE hard disk, P4 3GHz processor. When i try to install in graphics mode after disk checking i.e while anaconda is loading my monitor automatically turned off. So I recognized that problem with some driver. So I installed in text mode. Please tell me where do i get drivers for my system and how to install in my system. I don't have Internet connection for automatic installation using yum which was i found from linux forum so kindly tell me manual installation.
i installed the new beta 10.04 and it seemed right after the install and update that both nvidia hardware drivers were automatically install together. i deactivated both drivers. one driver showed the nvidia 173 driver and the other one showed "current" nvidia driver.
after a restart i then tried to activate the 173 driver. system required a restart. so i did. system booted to a black screen. i believe its at the desktop but i am unable to see it. i tried to hit esc at the boot screen to enter the grub menu but that didnt work.
First off, let me say: I'm a long time Linux hobbyist and recently installed Fedora 12. I've been looking (some say "lurking") around these forums for a bit after I loaded it up and let me say: Leigh, You're an amazing young man! Keep up the great work! I've a question: I've just put together a hobby computer: Pentuim 4, with 1 Gig of RAM and a Nvidia Geforce5700LE. When I install Fedora from a fresh install, everything works great: nice screen resolution, good graphics, etc. but no 3d or acceleration, so I'm very limited on my programs.
So I decided I need to upgrade my Nvidia driver. I followed Leigh's Nvidia setup guide after installing Fedora 12 (all four steps!) everything seemed to go fine, but when I reboot, I quickly see the Nvidia splash screen, but the screen then goes blank. Well, not really blank, I can see what appears to be a kind of cross-hatch pattern, plus I can see what should be my mouse cursor on screen (it also is a square shaped pattern). I suspect it may be at a screen resolution that my monitor doesn't support. Once there, the computer pretty much doesn't do anything else. Usually I can break into the x driver loading and go to the command prompt, but this latest time I can't even do that.
I've been googling such terms as "Gnome Shell Developer Reference", etc, and poking around developer.gnome.org; and i can't seem to find documentation on how to code interface elements against the gnome shell for javascript / xHTML / css. Does anyone know where they keep it, or am i completely misunderstanding that aspect of the new gnome hell, in that you only use js to do themes, but not plugins or inteface changes?
I have some trouble installing the nvidia driver on Fedora12. I started to follow the guide posted by Leigh here:[URL]..I enabled the rpmfusion repo, but when i wanted to install kmod-nvidia and xorg-x11-drv-nvidia-libs.i686 i had a depadency failure Here's what i got:
