Slashdot Mirror
NVIDIA Releases Optimus Linux Driver With New Features
An anonymous reader writes "Nearly one year after Linux creator Linux Torvalds publicly bashed NVIDIA and several years after their multi-GPU mobile technology premiered, the graphics vendor has finally delivered an Optimus-supported Linux driver. NVIDIA released the 319.12 Beta Linux driver that brings support for 'RandR 1.4 GPU provider objects' that basically allows for Optimus-like functionality when using the latest X Server, Linux kernel, and XRandR. The 319.12 beta also has many other features including better UEFI support, installer improvements, new pages on their settings panel, and new GPU support."
It's cool to see Linux gaming getting more attention.
GOOD JOB!!!
I can't get it working with the 3.8 kernel in the new ubuntu beta... wonder if this will make that project unnecessary..?
This is good news. I bought a new Intel/nvidia rig a few days ago and am now looking even more forward to using Linux on it! :)
So does this release bring the Linux drivers into parity with the Windows drivers? I'm sure this is a large step in the right direction, but if the Windows driver is still more capable or efficient, then Linux will still suffer on the gaming front.
"If a nation expects to be ignorant and free in a state of civilization, it expects what never was and never will be."
for as long as I can remember, and that is long
(Linuxer since 1991).
Never bought anything else for a display card though.
Explain that.
I love this picture of our fearless leader. Doing what we've all wanted to do to companies that fuck with us.
http://www.phoronix.net/image.php?id=0x2012&image=linus_nvidia_finger_med
ayottesoftware.com
it has nothing to do with gaming performance.
Of course it has to do with gaming performance. If you can't switch between the IGP and a discrete GPU without a reboot, then the launch and shutdown time for any high-performance 3D game includes a reboot to GPU mode, then a reboot to integrated graphics to save battery.
A thoughtful, complete, and informative reply on Slashdot. It's like 1998 or something! Thanks for that!
I noticed that it was doing that passthrough to the intel as well, but figured that might suffice... esp if it's something that can be turned on and off on-the-fly as needed (the way it apparently is on Windows...)
Anyway, hopefully bumblebee will still work w/it once I figure how to get it going with this new ubuntu 13.04 kernel.
Wonder if they've fixed the wonky error message about RandR missing when Xinerama is enabled yet...
I'll be glad when this is actually able to run on Lenovo's notebooks, which require an ugly ACPI hack to enable the Nvidia GPU: https://github.com/Bumblebee-Project/bbswitch/issues/2#issuecomment-3797568
I deny that I have not avoided attaining the opposite of that which I do not want.
Please NVIDIA do something about reliability, compatibility, provide debug symbols, meaningful error messages, and a way to easily provide feedback and response and the understanding of how the collected data is used rather than the impression it goes to /dev/null.
You have subtly reassigned your user base to serve as your beta test annoyance discovery team, selling hardware with drivers that provide the air of functionality but each with its own nuances of failure and glitches.
I try not to be nasty, but Linus's response was correct. It's time to draw the line and make up for the last 4 or 5 years of failed promises.
So what do "real gamers" (as you define them) do instead of gaming while riding the bus, train, or carpool to and from work?
GPL?
In the summary. Seemed funny to me.
haven't purchased anything (for myself or clients) with an nvidia chip in it for at least the last year. nvidia had time to design their way out of old third party impediments to open sourcing the driver code and they haven't even started. i don't care what their reasons are. I'm not installing their closed source (security and stability issues) code into a perfectly good linux machine and i don't appreciate their cavalier attitude towards me and mine as a market. The open source radeon driver (http://www.x.org/wiki/radeon) works really well these days on supported cards and i hope the rest of the community will vote with their wallet and send a message. AMD needs to double down while they have the chance. @nvidia: you think this whole linux thang is going away? You'll get yours...
nVidia has reasons for doing things the way they do. Yes, one of them is probably "because we don't want AMD grabbing our work," However there is some validity to that in that it is expensive to have a team of highly qualified people to do your development.
However that aside, there are licensing issues that keep their drivers closed, and there may be good reasons to want to use that code rather than try to re-implement it. Likewise there may be reasons to do their own thing and bypass some of the standard way of interfacing.
nVidia produces Linux drivers that work. They support the latest OpenGL features the hardware can handle, they are fast, and they are stable. That's pretty damn useful. So they are doing something right in their development. People should consider that, rather than just assuming that nVidia could easily deliver everything the same, but just in a format that makes OSS heads happy.
Also consider that maybe working with someone is an easier way to get at least some of what you want than fighting with them.
Is Linux users have been whining that nVidia should open up their drivers. nVidia won't, so LInux users thing nVidia is the bad guy and "jerking them around" rather than investigating if there might be some valid reasons.
However despite that ideological point, he still uses their products, because they are the best for Linux. That again is a reason I say maybe people should consider that nVidia has reasons behind what they do.
Who's this Linux Torvalds guy?
Somebody get Soulxkill his coffee.
As far as I know the only way to overclock with the binary driver, being as nVidia has neglected to implement clock control in the Linux driver for every GPU chipset from Fermi onwards, is still to set a static overclock in a modified firmware. In the role of a workstation this probably isn't an interesting feature to many users, but when you are comparing gaming performance the amount of additional 'free' performance you can extract from a given card alters where the available options rank in price/performance and is at least a consideration for many power users. The ability to alter the ratio of core to memory clock to reduce power consumption while optimizing performance for GPGPU tasks is also a consideration. I personally have purchased nVidia cards for several generations SOLELY on the basis of the stability, maturity, and feature-completeness of their drivers. After three iterations waiting for feature parity of a fairly essential function I am starting to seriously consider giving AMD's offerings another trial.
That's the real problem with nvidia.
but its too late for me, nvidia already lost me as a customer, i wont buy their products anymore and when shopping for a new laptop or desktop i always look for ATI video now, (i dont like having my PC half_broken because some snooty hardware MFG wont build decent Linux drivers
Politics is Treachery, Religion is Brainwashing
As far as I know, it doesn't exactly provide "linux support", but rather "Linux X86/X86-64 support", all numerous other platforms staying unsupported as always.
I've always understood THAT being the major complaint about their "linux support".
And of course in the future when binary compatibility will be broken for whatever reasons, guess if they are going to update their older binary-only drivers?
Not true, K3000M and up you can alter clock via nvidia-smi.
For most Linux laptop owners with less than 17" screen/chassis (where K3000M+ is not available) this effectively means you can either eat sh*t (i.e. live with stock clock speeds), or roll the dice with a vbios mod.
I have mixed feelings for Nvidia: on the one hand they do provide a Linux driver, one that is generally quite stable (in my experience) and feature rich (relative to alternative chips), so am thankful to Nvidia. On the other hand they fix clock speeds for mid-to-low end chips, and force max power mode with multi-head setups, thus pointlessly heating up the GPU and making laptop fans go on & off, on & off, on & off for no reason, and for this I have a finger for Nvidia.
Had to resort to vbios mod to undervolt/underclock the Nvidia chip in order to have some peace & quiet. Hopefully once latest driver is stable vbios mods won't be necessary anymore (i.e. integrated chip can drive the laptop and dedicated chip can driver the external in a simple 2 monitor setup).
But do people buy the kinds of games that run in DOSBox anymore? One Slashdot regular has repeatedly told me that developers of new games with retro style 2D graphics are living in the past.
I could overclock (and get fan speed control) just by setting Option Coolbits "5" in xorg.conf, then a new section appears in nvidia-settings. (the coolbits number is a binary mask about enabling three features so it can go from 0 to 7).
I learnt of this by finally reading the nvidia driver documentation, which was quite detailed and allowed me to learn the xorg option to bypass monitor EDID. Then I quickly disabled overclocking before of concern for stability - my card is an old 7600GT and I think it crashed on a 20MHz oveclock.
Just wonder if they are leaning toward mobile market as well. It makes sense if they want to target the tablet market, as there are more mobile and social games around these days. And thinking about the developing countries, there are huge market on the mobile/tablet market.
So that I can watch bluray movies on my PC
What makes paid streaming or paid downloads unacceptable?
and so that I can burn movies to watch on other things
Then buy "other things" that have a USB port, and load movies onto an external drive.
Read this *before* you experiment the drivers. https://plus.google.com/u/0/102207276811032054708/posts/8bAKax1PJoi New nVidia beta drivers, 319.12, have been released yesterday. Unfortunately, several web sites have been quick to put articles with titles suggesting Optimus support finally coming to Linux, and I'm saying "unfortunately" because I believe this is a case where inaccurate reporting hurts everyone: to non-technical users, the articles may have an effect of giving a false impression that the wait is over and the complete and proper support in the official binary drivers has arrived. As far as I see, there is some confusion and misinformation in users' discussions, and that is not helping either. So, let's try to clarify things. In short, the new beta is but a first user-visible step towards complete Optimus support. Remarkably, it covers use cases that Bumblebee has never supported well: using external monitors attached to the nVidia GPU, and running all rendering on the nVidia GPU. On the other hand, Bumblebee provided power management and render offloading on the basis of individual applications, neither of which is offered by the new beta. In a typical muxless Optimus laptop (or a mux'ed laptop in "Optimus" configuration), you have the laptop LCD panel connected to Intel GPU only (so that nVidia card cannot display on it), and you may also have some external video port connected to nVidia GPU only (so that Intel card cannot display on it). Normally, you run the X server with the Intel driver, with the only output being the LCD panel, and all works well. Let's now consider more fancy scenarios. 1. You want to run a graphically intensive game, so you wish that heavy rendering is performed by the nVidia card, but it's only powered on for the duration of the game. You don't need to redirect rendering of any other applications to the nVidia card. This is called "render offloading". 2. You want to temporarily plug in an external monitor into the nVidia-driven output port without disrupting your existing X session already running on the X server with the Intel driver. Since the Intel chip cannot access that external port, it will need the nVidia card to perform display ("scanout") for it (this is assuming Intel does all rendering; alternatively, nVidia could be performing rendering for its portion of the desktop). 3. You want to use nVidia card for rendering the whole desktop, trading increased power consumption for improved acceleration of all graphical apps, including the compositor. Since the nVidia chip cannot access the laptop LCD panel, it will need the Intel card to perform scanout for it. 2 and 3 are called scanout offloading, and notice how it is needed in different directions for different use cases. The card performing the scanout is called the scanout sink, and the other is called the scanout source. With the new beta, nVidia supports scanout offloading, with the restriction that the nVidia chip can be the scanout source but not the sink. Thus, it supports use case 3. Use case 2 needs GPU hotplug in the X server, because you want to power up the discrete GPU only when the external monitor is plugged in, and on top of that use case 1 needs a mechanism to route rendering between different drivers. For now, it's possible to use a combination of virtual crtc patch and hybrid-screenclone to "solve" case 2 (yep, that's painful), and Bumblebee "solves" case 1. Proper support in the drivers/server stack will be more efficient, of course. Notably, it should be possible to use the new beta drivers to get better accelerated rendering for gaming sessions by starting the game on a separate X display with nVidia driver and scanout offloading. FSGamer should come in handy for that. Note that you want to be using xf86-video-intel driver for the offload sink in this case (not the modesetting driver as the readme currently suggests), which should work since version 2.21.5 and required, since Xorg does not support different drive
For a long time, I have decided to boycott Nvidia (which I have nicknamed as “Hang-vidia” due to the fact their drivers frequently caused my machine to hang) due to their positive hostility for Linux, and open source, and what not (lack of support for open source efforts, no specifications released, legal threats against open source efforts, dropping support for old cards, etc.), and the low quality of their binary-only offerings (frequent hangs and crashes), and their general incompetence. I will never buy Nvidia until they release SPECs and make their driver open source. See my old petition about that.
After using an old GeForce 4 card where neither the "nv" driver nor the "nvidia" driver worked properly, I switched to an ATI Radeon HD 2600 Pro card, and it served me extremely well, and was rock solid. Now I have the built-in Intel graphics on this Core i3 machine, which causes some problems, so I may opt to buy a new (and probably better) ATI/AMD card. But I'd rather be hanged than buy hang-vidia.
We have two eyes and ten fingers so we will type five times as much as we read. http://www.shlomifish.org/
He should drop that F and the finger at Microsoft, if it works for Nvidia it will work for Microsoft.