Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
Graphics Software

Tom's Hardware Linux NVidia Benchmarks 156

diehard writes: "Tom's Hardware has posted a set of benchmarks of NVidia cards running under Xfree86 4. They are pretty impressive - it looks like Linux has finally become truly viable for gaming."
This discussion has been archived. No new comments can be posted.

Tom's Hardware Linux NVidia Benchmarks

Comments Filter:
  • by Anonymous Coward
    *rolls eyes* Troll

    Do some research. Nvidia clearly said IN ABOUT 20 INTERVIEWS that they'd liek to Open Source their drivers *BUT* they have signed various NDAs on sections of their technology, and they would be in direct violation if they released the source.

    I've had lunch with one of the nvidia linux developers, and he laid the whole thing out. Believe all the conspiracies you want, I don't care, but don't go spreading FUD.
  • logitech firstmouse+ (with wheel), using the "IMPS/2" protocol. here are the relevant xf86config lines:
    Driver "mouse"
    Option "Protocol" "IMPS/2"
    Option "Device" "/dev/mouse"
    Option "ZAxisMapping" "4 5"
    Option "Resolution" "1000"
    Option "SampleRate" "85"

    I also have "xset m 0 0" on.

    I've tried all sorts of variations on these numbers, I've tried booting X with no window manager and a single xterm, etc etc etc.

    As I said, my mouse performance in windows is excellent, so hardware shouldn't be the problem.

  • Standard Linux Troll Response #41:

    Linux as an operating system can and should handle all of the tasks we're asking of a desktop computer today. The mainstream, easy to use OS will help me play UT (have you heard about the unappreciated art form that is FPS?) Don't count out games as one of the driving (pun intended:) forces behind the growing demand for big iron power in little iron boxes. I see little danger in Linux becoming a "gaming" OS.

    I'd hate to see a bunch of punkass newbie kids come in and rune the powerful thinking man's little slice of heaven ;j
  • As far as I know, the NVidia-drivers are
    closed source.
    Has something changed the last 30 days?
  • Yep. i'm Raul. Maybe one of these days we'll be slashdotized and get a few thoushand hits, you listening CmdrTaco? :-) Thanks Flo.
  • SGI's transposition to IA-32 has got to be one of the worst mistakes in the company's history. SGI Linux boxes are not worth the exorbant amount of money they cost consumers. You can get IA-32 workstations from almost anyone that support PCI 2.1 and UDMA standards. The NUMA system used in the MIPS based Origin boxes was something thats pretty limited to MIPS. Their Intel machines are no more powerful than anyone else's Intel machines, the only real difference is the price. Please stop bringing SGI into these sorts of conversations. Their adoption of Windows NT was a PR stunt as was their adoption of Linux. The only reason they ported Linux to their Origins was to give a little life into a product they're no longer working on improving. An Origin running Linux won't be any faster than it was running Irix but it will have a larger library of programs that will run on it with a minimum of tweaking. No one is really developing for Irix anymore and SGI knows that operating systems don't mean nearly as much as the actual applications do. They're now merely opening up and/or porting their software and such so whoever ends up buying them won't own some important pieces of software like Maya and OpenGL.
  • Pick either:
    1. NVidia hasn't created a proper install-script for it (NVidias fault, not an inherit fault of Linux).
    2. The drivers are not opensource, so many distributors hesitate in distributing the NVidia driver with Linux, making point 1 obsolete. Again not an inherit fault of Linux. NVidia can remedy this as well.

    I'm not trying to be a troll, but it is quite annoying when some people give Linux and XFree86
    the blame for something that is in the hands of
    the manufacturer (NVidia).
    That said, I'll probably buy an ATI radeon, because of the opensource-nature of the drivers.
    Why does NVidia has so much to hide, when ATI doesn't?
  • I don't recall the details, cause I *still* haven't had time to mess with Linux Q3A, but a friend of mine said something about renice-ing X to fix the mouse performance issues. He also mentioned a script that will do it for you (among other things) before starting up Quake. Sorry for being so vague, but there *is* a solution out there.
  • > Is there ANY graphics card in the same class (or better) as NVidia

    No. The Voodoo5 needs to be plugged into a hard disk cable (in the V5 6000, the mains - whoever heard of a video card you had to plug into a wall?) and still gets beaten by the GTS (and even the GeForce DDR) in benchmarks, the Matrox G400 [although nice] is a few generations old (it's TNT2-Ultra generation), and... what's left? S3 makes budget chips, PowerVR's no longer a factor (or has drivers)... nVidia's the best we've got!

    If you want something in a lesser class than a nVidia card (and I'm just thankful that there are drivers, whether they're OSS or not) you might like to head towards the Matrox G400, which is a Nice Card. But not a GeForce GTS, which is hopefully going to be my next video unit. Until the NV20 comes along...
  • Can you count? In order to make things a little more palitable for those who were not trained using SysV SuSE decided to go numerologically. 0 is halt, S for single user mode, 1 for multi-user no network, 2 for multi-user and network, 3 for multi-user with network and XDM starting at boot with 6 being reboot. 1, 2, and of course 3 for the main init modes people are going to be using. SysV wasn't designed for uninitiated users while SuSE's market is a gaggle of uninitiated users.
  • Has somone ported the little tiles you get with the DOS version yet? They rock, & I didn't see them mentioned at all for the Linux Version.


    --

  • yes -- driver model -- can't hack it
  • I would have to disagree with most of what you have said.
    SGI going to IA32/IA64 is an obvious choice.
    The MIPS is falling behind in mhz.
    SGI shops feel comfortable staying with the SGI name but they will leave them if the price/performance ratio drops too much.

    I agree that the SGI web servers are a joke... they are manufactured by VA, I believe, so why not buy a VA box?

    I also agree with the switch to NT. But I also know that the people I work with that want the SGI NT boxes will not give them up for Octanes.

    Obviously you haven't been paying attention to SGI press releases... The Origin line has gone from the 200 to the 2000 to the 3000 series, all in the last year or two. Each of them still being sold and produced. They are expanding the line.
    By IA64 compatability in the 3000 line, they are allowing themselves to take advantage of a largely distributed CPU that will continue to be developed and improved.

    Also... most programs that run under Linux are also ported for Irix. And... even if the Origins do run Linux on the MIPS, the code still needs to be ported for the chip, so that argument is moot.

    Maya is still being developed for the SGI, so people are still developing for it. Maya is a killer app. It is why we have 50+ SGI boxes in house. I know other studios are the same.

    As I have said, people buy SGI for the name. The name implies a level of quality and support. Big places spend money for products they trust.

    Don't go counting out SGI yet.
  • by Anonymous Coward
    do you really sit in front of your computer for an hour doing nothing but watching the progress bar move? Why do I find it hard to believe that anyone pays you 15 dollars an hour, much less 50?

    Your highschool level knowledge of economics is better than most slashdotters, I'll give you that.
  • I guess it also has something to do with the type of movement in a FPS compated to tv or film.

    In tv/film you mostly have very similar images following each other, e.g. a person walks across the screen, but the background remains the same. In this kind of a situation, at 25fps your brains 'blurs' the images together and everything appears to be fluid.

    In FPS games on the other hand, it is common for 2 consecutive images to be completely different, e.g. when you do a 180 mouse flick to look behind you. In maybe 0.5s the background changes completely. In this kind of situation you brain might actually notice that 25fps, while 60fps would be much more comfortable.
  • The xfree86 FAQ at their site says "If you want 640x480 without scrolling, remove all the larger modes." This is not acceptable.
  • You can patch directX 5 into NT4.

    Email me nicely using (and damn well use pgp!) at teo@nvnetworking.com and maybe i'll send you a copy.

  • I have many times gone on there.. ripperda is no help at all.. Hes asked me to email him, and I have several times.. again, he doesnt respond. I guess my website aint big enough to garner tech support. And just last night, I msged him again with my problem, his response was that hes never heard of that problem happening (so either he doesnt read my emails, or nvidias mail server is as flakey as their drivers) and to email him again.. So I have.. again, no response.. The users on #nvidia are somewhat ok, but usually go WAY off topic WAY too much and your question(s) end up getting mixed up in a conversation in which classic computer was better, the c64 or the atari?
  • Go onto irc.openprojects.net, #nvidia. I found this to be an excellent source of technical support for my problems. They have nVidia staff online to help you out.
  • Huh? Could you please elaborate?
  • The big problem is getting graphics chip specs from graphics chip makers.

    Dirk Hohndel (XFree core team member and SuSE CTO) said in his speach at LinuxTag that it has become MUCH better, both because Linux is getting more important and because the gfx card makers get more confidence in the word of the XFree team not to break NDAs etc. He gave the example of Matrox. IIRC, for the first card they didnt give any documentation, for the second some after it had shipped and now (for the G400?) they gave complete documentation before it shipped.
  • I love the comment on toms site on how nvidia is working with him to fix out the bugs. Funny how nvidia doesnt do that with end users. I know of at least 4 other users who have emailed nvidia about problems with the linux drivers/gl drivers and nvidia doesnt even care to respond. They have no bugzilla setup, no nothing to track bugs or whether or not they even received your email. (and man, there are some huge bugs in nvidia drivers, from console switch lock ups to X not even loading) I guess I have to own my own website before they will respond to me.. oh well.
  • by molo ( 94384 ) on Saturday August 12, 2000 @10:50AM (#860187) Journal
    Tom has taken a big step for him in dedicating a review to Linux. He has the following to say about it:

    This was my first Linux hardware review and it will certainly not be my last. The first time is always supposed to be the hardest, but the most rewarding as well. Please let me know how I performed here. Was I babbling too much about Linux? Weren't there enough facts in the review? Don't you care about 3D stuff in your Linux-box? Please give me feed back under tomslinux@tomshardware.com [mailto]. I will try to live up to the expectations of the Linux community, but first I need to know what they are.

    We need to encourage him! Tell him some of the things we would like reviewed with linux benchmarks. Thank him for taking a big step in dedicating a whole review to linux. Only good things can come of this!
  • Try a USB mouse. My Microsoft Intellimouse refreshes 80 times a second under XFree4.
  • You need DGA to be activated...is it?

  • Unfortunately, Linux had lower benchmark scores on nearly every test, at nearly every resolution. Marginally, of course, but it's still there.

    I would go for Linux as my primary gaming platform if they were more games (like Deus Ex, The Sims, etc.) and more support for games. As it stands, Win2000 is my primary gaming platform, with Linux as my programming platform.

  • ... you might want to avoid booting right into the GUI. Make sure that you change the line 'id:3:initdefault' to 'id:2:initdefault' in the file '/etc/inittab'.


    Now, maybe I'm an idiot, but I could have sworn that runlevel 5 is generally the one that runs XDM/GDM/etc by default. When I 'init 2' I get a system with no NFS, and no eth0!

    Just a thought.
  • With the next release of RedHat it should autodetect your video and with any luck automatically install the correct configuration for 3D acceleration.
  • by iCEBaLM ( 34905 ) on Saturday August 12, 2000 @11:10AM (#860193)
    You might wish to try the IRC channel #nvidia on irc.openprojects.net as we have very experienced users there that can help you with most installation problems, as well one of the nvidia linux developers is a regular. We often get between release "experimental" drivers which fix some bugs and help with final releases, as well as a set of pentiumpro+ optimized drivers which seem to boost 2D performance significantly and 3D performance somewhat, YMMV.

    There are some pretty bad bugs in the current drivers which were fixed previously but crept back in somehow.

    -- iCEBaLM
  • explain unpowered PS2/VGA switchboxs then?
  • "1) There are no HOWTO's or webpages dedicated to your card. When you ask a simple question, you will immediately be bombarded by three morons who say "why didn't you read through the last 712 messages! everything you need is there!""

    Actually, everything you should need is really on these two pages:

    http://dri.sourceforge.net/DRIcompile.html

    and

    http://dri.sourceforge.net/DRIuserguide.html

    Ranessin
  • Its good to see a major site like Tom's Hardware giving Linux some limelight.
    As for those 'l33t users' who say that games will somehow damage your precious OS, this is nonsense. Thats the wonderful thing about Linux, because its Open Source, and there are many distros, you can tune it to do exactly what you want (unlike with other OSs); you want a console-only network station for a l33t user, get Slackware or Debian, or grow your own distro. You want games, get SuSe, or Redhat or something and install high powered drivers, and games and stuff; ITS YOUR CHOICE; YOUR POWER! USE IT!
  • explain unpowered PS2/VGA switchboxs then?

    Okay.

    I have an unpowered KVM switchbox. With some of my systems it worked. Other systems would lock solid or just lose keyboard or mouse support until rebooted. I've since purchased a decent, inexpensive powered KVM, and the problems have gone away.

  • In one respect, I agree with you.
    I too wouldn't want to see Linux become
    what Win98 is today, more of the main
    platform for games than for productivity.

    It irks me that some people try to get into
    Linux not because of the advantages of using
    the OS but because it's 'cool' and its all
    hype these days. I had a relative that tried
    to mimic me and set up a linux server,
    which I'm happy to say that he failed
    miserably.

    Why did he fail? It wasn't because he
    didn't have the resources available to him
    that I did? no. It wasn't because he had
    inferior hardware to me? quite the contrary.
    He failed because he didn't take it seriously.
    All he was concerned about was feeding his
    ego. And in that respect, I agree with your
    statement. If you're not going to bother
    to get serious with Linux, then stay away.
    Because the last thing anyone wants is for
    disgruntled users who've tried and failed to
    use linux trash it to everyone else.

    IMHO,
  • No, OpenGL/GLX doesn't give specifications of the card's capabilities (At least last time I looked). Like for example, I might want to enable bump mapping for better quility, but if the card doesn't support that in hardware, I'd prefer to disable it for the speed.

    Even if you have an API like DirectX which tells you the capabilities of the hardware, to do it right you have to ignore this and just time a frame with bump mapping and then decide whether to use it. For example bump mapping might only run in hardware, if you have less than 3 lights. How do you want to describe this with capability bits (flags). Also, even if it runs in hardware, but is much slower than an alternative (say, using many polys) that looks as good, then you would use the alternative. As a gfx developer you are not interested in what runs in hardware, but in how fast it is. Also, like the other person wrote you can leave these decisions to the user, since its also a matter of taste.
  • Do you mean that there now exists an agreed upon standard for development such as OpenGL?
  • Comment removed based on user account deletion
  • Yes. Splendid. I'm the stupid quake gamer. Right. Can you also come up with some good arguments for the stupid quake gamer why gaming does not belong on a linux platform, if it is nevertheless quite possible to do so?

    I mean, it's not like we're tapping into your cpu cycles or anything.. Why are you so scared by drivers that might come with a standard distribution that happen to cover the leading 2D/3D graphics board in consumer pc's, and that users can choose not to use. I'm sorry but this doesn't make any sense. Keeping windows alongside should in the end not be a necessity, but you make it sound like it should be.

    Ohyeah, FYI (probably not), only Quake I and Unreal run acceptably on my PPro 180 Banshee (that's right, I don't even *have* an nVidia card) System. Gnome, KDE & Enlightment however are a totally different set of tears, and that's just static bitmaps!

    Over to you, Frank.
  • You're certainly the one person whose graphical opinions I trust. However, I personally find the tearing associated with blts to be visually distracting. It's not *so* bad in Quake engine games, but is quite obvious in others. Would the following be a valid solution? Composite your image in another buffer. Then blt from the composite buffer into the linear back buffer and when the blt is complete, swap the front and back buffers on the next vsync? I feel that many people are starting to ignore tearing just because they want the highest possible framerate and it concerns me. Tearing is an artifact that ought to be solved; we should not have to learn to live with it, IMHO.
  • by 91degrees ( 207121 ) on Saturday August 12, 2000 @10:20AM (#860204) Journal
    Well, this is a good start. Perhaps now we can add some of the other things that gamers like - for example an ability to change resolution and depth when not running as root.

    And how about a 3D API that allow you to optimise speed for the capabilities of the graphics card. Or some low level support for 3D cards in the frame buffer device (which could solve both these problems quite easily)

    I think I'm the only one who prefered the old days when the OS was considered irrelevent for games, except for as a program launcher, and something that should be disposed of as soon as possible. How many people want to multitask their web server with quake anyway?
  • I mean, if a sane person read that, without the 'mutherfucker' and 'idiot' in the post, they probably would agree.
  • Most of the coments I see are gamers pondering if it is worth it yet to game on the Linux platform, even with the better benchmarks. Some seem to think that because the mAcro$oft becnhmarks are still a tad better that they should stick with the gamers OS of choice.
    I come from the other end of the spectrum. I am anxiously awaiting these near similar benchmarks. Linux graphics machines are going to be a reality soon. The A|W port of Maya is going to be ready next year, the SGI Linux boxes exist, and now, the drivers seem like they will be ready for the release.
    This is great news. All SGI shops now have an alternative to NT. I know that many are evaluating the choice of SGI vs. NT with the increased NT performance and the lower prices. With Linux in the mix, good things can happen.

    If you look at what SGI has been doing in the last couple of months, they have been setting themselves up for this switch.
    1) SGI has ported Linux to work on the Origin series of machines.
    2) SGI has designed the next series of Origins to run on either IA64 or mips (with Linux support)
    3) SGI has produced Linux graphics workstations with more standard hardware than the NT only boxes of a few years ago.
    4) SGI has worked closely with NVidia for graphics hardware solutions
    5) SGI has signed a deal with Intergraph (one of the leaders in NT graphics boxes) to sell Intergraph boxes

    The first four point to Linux in the graphics world. The last one seems to either be a failsafe or.. as I see it, a chance to sell more boxes that could be dual boot graphics stations with some research.

    The next year will be interesting for all places that are SGI shops.
  • So then you would be Raul Alonso, I guess? Or someone else who worked on this? Anyhow, I just wanted to say thanks to you people for a hell of a great program, and to everyone else who happens to have a decent OpenGL card, go check his link out and get OpenUniverse now.

    I used ssystem as a screensaver on my Voodoo back in the days and liked it a lot, so I was interested to find the recent OpenUniverse review in C'T magazine. And I have to say I'm very impressed with this program on my TNT2. Visually stunning as well as educational I've spent quite some hours zooming around the solar system already.

    Thanks,
    Flo

  • Fortunately, though, Debian (which is the only true Linux out there) still costs $0 to install/use from the 'net. And the CDs, from www.cheapbytes.com [cheapbytes.com] for instance, are little more than what blanks would cost ($7 US for two CDs).

    Since Debian has its roots in the GNU project, and is the only GNU distribution of Linux (and therefore entirely open-source/free), it would lead me to consider it the best choice out of all the Linux distros.

    Sarge
  • But the 3D API you mention would have to live somewhere. Where do you intend to stick it, since you want to get rid of the OS ?

    And stuff like memory management ? file I/O ?

    you also want to get back to the good old days of having raw file i/o for your programs, or is this all supposed to be standardized in hardware ?

  • i wholeheartedly agree. i would enjoy playing a game or 2 of GLquake, but it's just to much of a hassle to set it up for my banshee to be worth it.

    i think the distro's should come with the drivers, and on installation should setup the proper drivers for playing games for you video card. and should also provide a way to keep it current. a couple of rpms, mandrake update, or apt-get, or somthing of the sort.

    not only would this be great, and make gaming under linux, actually, just gaming. but it would attract new people to check it out under linux, and buy linux games, which would result in more ports. and lastly, this would be even better than the system windows currently uses ;) so it would not just be matching windows in some respect, but beating it to somthing :)

    -------

  • Crap.

    I think basicly any linux app still has a long way to go before they become usable on the desktop. And while applications are what make OS'es work or not, Games considerably help push the numbers up in a steady pace, both in hardware and software.

    What harm will it do to finally have some drivers you can choose to install ? If there is a (large) demand for such things, and the people even get what they want, for free, then why does the much celebrated linux diehard guru has to have a problem with gaming all of a suddden?

    Man, I don't mean to be rude, but Linux is slowly changing shape if you hadn't noticed. Linux strives to compare with (and surpass) windows. Well, it better cover every inch of the playing field (read: market), not just that of the skilled & savvy power users, or someone is going to pull the plugg on his development cycle. Does this mean my aunty has to able to run linux ? maybe.. does that mean power users will have to button-click ? I hope not.. should we all be free to do as we like ?

    ...ah.. this trick question, I'll leave to you, the power user.

  • Debian is NOT free. It costs way more than $0, since you or someone else has to pay for the 1GB+ that you intend to push down the 'net.

    So, until you give me a free ADSL or something or other, don't claim that downloading stuff is free.

    And even the "free ISPs" require you to spend some money, in the form of hardware or other

    Math question: How many cds can you order from cheapbytes before it makes sense to buy the modem ?

  • nor attempt in any other manner to obtain the source code.

    So adovocating for open source (as a general pricipal or for this driver in particular) would put you in violation of the agreement ... and nvdia's lawyers will come and take your 1st born child (of graphics card whichever is more precious to you :-)

  • Well, this is a good start. Perhaps now we can add some of the other things that gamers like - for example an ability to change resolution and depth when not running as root.

    Resolution switching is already there, color depth is a problem, I agree.

    And how about a 3D API that allow you to optimise speed for the capabilities of the graphics card.

    Like, for instance, OpenGL/GLX?

    Or some low level support for 3D cards in the frame buffer device (which could solve both these problems quite easily)

    Which wouldn't be dissimilar to DRI's DRM?

    The architecture exists already, everything that is needed for great 3D in Linux is already there, it just needs to be USED and IMPLIMENTED. The big problem is getting graphics chip specs from graphics chip makers.

    -- iCEBaLM
  • A) It doesn't screw everything else. Everything else just doesn't get very many resources. And, it only happens while you're running the game. How often have you played Quake, and done something else at the same time? A lot of "hog" features are dependant on what the app asks for. If it is a simple game, that one could play in a window while doing something else, then that app can ask for a "cooperation level" that allows it to coexist peacefully with other apps. However, if it is a demanding game like Quake, DirectX allows it to use all the resources of the system. However, since 99% of the people can't play Quake and do something else at the same time, it doesn't matter. The minute you pause and switch out of Quake, then the resources go back to the applications you're using.

    The whole OS doesn't run in user-mode, just services that don't need Ring0. And a lot of services don't. For example, if you're drawing a line, you don't need to go into ring 0 to do it. Or if you're asking for information about a font. Remember, Windows is essentially, Linux +X+GNOME+Mozilla. In Linux, X,GNOME and Mozilla run in user-space anyway.
  • I have no clue what PGP is. And I've never heard of a way to patch DirectX 5 into NT4. The HAL won't let you do it. You can get DirectX 6 to appear as the installed version, but that's only because DirectShow or something runs up to version 6 on NT, and the DirectX control panel reports the highest component as the installed version. How exactly is this done?
  • I think Linus is wrong on this issue. (Yea, I'm arguing with god, shoot me.) Dependance on source compatiblitiy is a BAD thing. It leads to lack of cleary defined interfaces. It leads to non-replacibility, and all sorts of ugly things. Plus, it practice, it is not necesarily the best thing. Take a look at Windows drivers. 99% of them are closed source. Many of them are superior to their Linux counterparts. If having a stable driver API didn't work, why are Windows drivers stable? Maybe it's just me, I like COM, I like C++, so maybe I don't "get it." However, I point to BeOS as a good example of a driver API. They have dynamically loading, binary drivers. However, they maintain API versions. Thus, if an API break absolutely needs to occur, all changes can be lumped together and released as a new driver API. Now the kernel will still be able to load old drivers with a minimum of performance hit, and bloat really doesn't increase, because new driver APIs are rare. Above all, a stable driver API encourages stable code. Since the driver API is solid and non-changing, drivers can get maturity without having all types of changes. Face it, changing old source to use a new interface is an inherently dangerous process full of hacks and workarounds. Not a good idea for something as critical as a driver.
  • A) Texture RAM is cheap. For the last few years, graphics cards have doubled the amount of texture memory every year. Watching texture RAM is not something that people will have to do anytime soon.
    B) Memory bandwidth is a much bigger problem than the amount. However, this one will soon be solved as well. The PSX2 is already using embeeded video memory, and the GCube, (with 32MB or embedded memory) is what PC graphics cards will look like in a few years. It would not be hard to imagine a card with a large bank of main memory, a large texture cache, and compression features to move compressed texture over the memory bus into the texture cache. Or, they might just use freakishly fast RAM like Hercules is doing on their cards. Or go the Matrox route and use very wide busses. There are many solutions to the problem.
  • you stole my sig you bastard.
  • I dont think anyone's claiming 3d on linux is easy, at least not nearly as easy as setting up your card on Windoze, but it is certainly NOT "weak".

    I have a P2/400 and a Creative 3d blaster annhiliator 2 (forget the exact name.. its the caard with the Geforce2 GTS) and can play q3 at 1600x1200 with everything set up to the max. Runs very smooth. No mouse problems either.

    And that was with the standard 0.9-4 drivers from nvidia's website. I went to irc.openprojects.net and talked to the guys in #nvidia and got drivers optimized for the PPro (or better) and got a 5-10% bump in 3d performance. (Using the mesa "gears" demo to measure)

    siri

  • Having the source to the driver means that if I'm sufficiently clued, I can make the driver work with my Linux 7.2 system. If I'm not sufficiently clued, then I can learn. Not having the source means that I'll have to figure out how the entire card works, which is going to became sufficiently close to impossible in the near future that it's not worth it as cards become more complicated.

    The Linux driver API is never going to be stable. Linus has said this. Even when the external API remains stable, the internal kernel structures may change. A module that compiles on both 2.2.12 and 2.2.16 may not work if I try to install one compiled under 2.2.12 on a 2.2.16 system. At the end of the day, the Linux kernel is open source. Anything that interfaces with the kernel and is not open source is going to increase the difficulty for users and is significantly more likely to cause problems.
  • NT actually performs better in Quake than Win98 does.

    Oh? On my machine software-rendered Quake was doing 50% of performance on NT compared to 98. So...
  • Hmmm....

    For some time I've been thinking of preparing a 'sub-distribution' (i.e., take an existing distribution, reconfigure and repackage) targeted at the 'Gametop' (gameheads whose main purpose for using a computer is to play games). It would look something like:

    GAMELINUX

    • Based on Debian (because that's my favorite distribution).
    • Hosted on an UMSDOS filesystem. Once the gamehead graduates from being a newbie and becomes more proficient he can start using ext2fs but it seems beter not to bother with partitioning at the beginning.
    • Distributed in a .zip file (zipslack-style). Easiest setup I've ever seen to get the basic system running.
    • Including Vampire, a tool that will suck config info (e.g., video card, monitor) from the Win9x registry and write it to the corresponding configuration files in Linux (the user can preview the results and select what gets written).
    • With LOTS of free games to select from (e.g., nethack).
    • With an optional menu system so that the gamehead can get to the game without learning any command.

    The conquest of the Gametop is a necessary step in the quest for world domination. The ingredients are available, we just need to put them together.

    Various Legal notices:

    • If you think that Linux is already available in too many distributions, please ignore this post.
    • If you like these ideas I propose that we work together on them. If instead you choose to use them without my participation you MUST make the resulting product free in the following 2 ways:
      1. It must be Open Source (as in free speech)
      2. If you sell copies, you must provide, free of charge, at least 1 copy of each released product to me (as in free beer :)

    Enjoy,

    Eusebio (@corelcity.com)

  • You're wrong. I hate to start a post that way, but if you think you have your facts correct, then post a URL to either a Microsoft or Intel web page (or any other manufacturer) that states that you are correct. Plugging a PS/2 peripheral, keyboard or mouse, while the system is running may cause the motherboard to stop responding. The OS you run doesn't even enter into it. My friend, Justin, has also fried his motherboard's PS/2 controller by plugging in a PS/2 mouse that had become loose. The MSI motherboard and Logitech mouse both bear the Windows 9x logo.
  • On my machine hardware rendered Quake 20% faster and a lot smoother (consitancy of framerate) than on Win98. On NT, my OpenGL apps (again hardware rendered) run about 15% faster. So... The truth is that in general, NT performs better for OpenGL applications. Notice I said OpenGL applications. I'm pretty sure Quake's software renderer doesn't use OpenGL. Either way, it doesn't matter what your particular machine does. In general concensus is that NT performs better for OpenGL.
  • by 1%warren ( 78514 ) <<wardon> <at> <xtra.co.nz>> on Saturday August 12, 2000 @03:33PM (#860227) Homepage
    FYI, getting the DRI drivers to work with 4.01 is fairly easy (this supposes you have a V3 & not a V5)

    dload xfree86.org linux binaries of 4.01

    run xinstall.sh

    run XFree86 -configure

    edit your new XF86Config till it works properly (hint - do a xf86config & copy details from that to your "-configure" generated one - you need to select "generic vga" as your card though)

    get: tdfx_drm-1.0-2.src.rpm & Glide_V3-DRI-3.10-6.i386.rpm from: here [3dfx.com] & follow the instructions on the page to install them

    do a "modprobe tdfx" to load the module (before you "startx")

    Quake III should now work, including DGA mouse (YMMV)(hint - create an .xinitrc in ~/ with "exec quake3" as the only command - if you change res ingame it tends to fsck your dektop on exit)

    I agree that the documentation is sparse (nvidia does it much better), but it's early days for these drivers, & they are being actively developed.


    --

  • Not to beat a dead horse and/or start another religious war, but it seems to me that the main reason installing this stuff is so hard is that it isn't bundled with XFree. Why? Oh yeah, they won't release the source...
  • Maybe you didn't see that Tom was using SUSE (the picture at the start of the article). SUSE launches X at runlevel 3. Redhat uses 5. I think Debian uses 2.
  • I remember about a year ago, I got a very unofficial pack that sorta put dx5 (limited) onto NT4. It let a bunch of things run, but most others wouldn't... Not quite the real thing, but a good hack for some.

    PGP is Pretty Good Privacy. Encryption, digital signatures... good stuff. Check out www.pgp.com or http://www.pgpI.org/ if you are outside of the USA.

    --
  • by John Carmack ( 101025 ) on Saturday August 12, 2000 @03:59PM (#860236)
    All signs are pointing towards a future without page flipping, so adding the messy infrastructure for it now would be a mistake. Don't let benchmarking furor encourage a messy code architecture.

    Points:

    The benefit of page flipping is decreasing as more and more computation is done per pixel to the back buffer.

    In the old days of 2D scrollers, you might barely cover the screen with one pass of writes, so page flipping could double your speed over blitting.

    On a typical modern 3D game that becomes fill limited, under 25% of the performance is in the blit, and often under 10% in scenes with significant overdraw.

    In upcoming games that composite 20+ layers of textures, the cost of a blit is down in the noise.

    Blits add flexibility. Anti-aliasing is better done through a blit operation than with a deep front buffer. Other operations, like converting from a 64 bit work pixel to a 32 bit display pixel, or performing convolutions, are also better done with blits.

    Back buffers are more optimally arranged in tiled patterns, while front buffers prefer linear scans.

    Basically, our back buffers are starting to look less like raster

    Page flipping doesn't apply to windowed rendering unless you butcher the X server to render all 2D to multiple buffers and clip all 3D operations. I consider that a bad thing. Making the full screen rendering more distant from windowed rendering is also a bad thing.

    Every implementation of page flipping brings in a class of bugs, and obfuscates several code paths. It's not worth it.

    John Carmack
  • by Skald ( 140034 ) on Saturday August 12, 2000 @05:14PM (#860240)
    it looks like Linux has finally become truly viable for gaming

    I beg your pardon... Nethack has looked great on Linux for years!

  • I'm sticking myself out on a karma limb here, but what the hell. Some things are worth taking a hit for.

    The post I have replied to is NOT a troll, despite having been moderated as such. The poster has presented an entirely on-topic opinion (that, as an aside, I tend to agree with)

    You, as a moderator, may or may not agree with the post, but not sharing the same viewpoint does not entitle you to slap him with "Troll" - especially when there are so many other REAL trolls to slap down.

    I'll be looking for this one in meta-mod....

    Good post Mr. AC.

  • Just because NVidia may (or may not) privately express their wish to Open Source their drivers doesn't change the _real_ problem one whit.

    The problem is not some metaphysical "corruption" of a "pure and free" Linux system, it's a much more practical issue.

    The Linux kernel is under constant revision and continual development. Sometimes, the driver API changes - typically for a very good reason - and then any binary-only driver will break.

    This means that in order for your binary-only driver to continue to function as the kernel is revised, the provider of that binary-only driver must keep pace with the mainstream kernel development.

    Furthermore, in the case of stable-API but unstable behaviour, it is very difficult to get an almost-functioning binary-only driver fixed in any sort of reasonable timeframe, nor is it easy to debug instabilities that may arise through driver/kernel interaction.

    The only real solution to these technical problems is to release source.

    Let me put it this way - any company that releases binary-only drivers must needs become a slave to maintaining these drivers. If they aren't up to the task, then the result is poor or no support as the kernel mutates through successive versions.

    Binary-only drivers SUCK from a purly technical perspective. That's not FUD, it's truth.

  • A better word: Links


    He who knows not, and knows he knows not is a wise man
  • by ink ( 4325 ) on Saturday August 12, 2000 @11:45AM (#860251) Homepage
    It would be more interesting to see how Linux fares against Windows NT/2000 rather than Windows 98se. NT and Linux have much more in common than the DOS line does; and NT shares many of the high performance gaming problems that Linux has to address as well (like protected memory, premptive multitiasking, userland libraries and such). Window 98se's job is to get out of the way of games. NT and Linux still have to have running operating systems underneath.

    The wheel is turning but the hamster is dead.

  • by dirtyhank ( 135909 ) on Saturday August 12, 2000 @12:09PM (#860252) Homepage
    I'm the author of the proggy Tom used as one of the benchmarks, the unix solar system simulation as he called :-). He didn't put a link so just in case you want to give it a try you can found it right here [www1.las.es].

    However, i have to tell you this a rather old program and maybe not the most suitable for benchmarking a new generation 3D cards (like the GeForce).
  • by Straker Skunk ( 16970 ) on Saturday August 12, 2000 @11:46AM (#860253)
    VA is currently working on an XFree4.0 driver for the new Radeon cards. It's being developed under contract from ATI themselves, it'll use the DRI, and it'll be open-source.

    Check out the adjacent article on Tom's, about the Radeon. It performs slightly slower than the GeForce in 16-bit mode, and at lower resolutions, but at high-res 32-bit modes the Radeon edges out the GeForce.

    The GeForce numbers may be impressive, but me, I'm going to wait until the Radeon driver is ready. And then I'm going to show ATI just how much I appreciate an open solution.
  • by Kinlan ( 138030 ) <paul@switc h m e d i a . c o .uk> on Saturday August 12, 2000 @12:11PM (#860254) Homepage
    Really,

    Linux has a long way to go to being truly viable as a gaming platform. Just becasue one (or a few) video cards have good bench marks, doesn't mean that it is viable, to be viable there needs to be GOOD support from all vendors, who specifically target Linux as a gaming platform like they do to windows, not just one or two companies.

    > Most gamers don't have the latest hardware or the fastest as it is normally very expensive.
    We need excellent support from all hardware manufacturers that have produced video cards over the last year or two. So that all linux users can enjoy a great gaming experience.

    This is just the tip of the iceburg, we also need better support for mice, sound cards etc, I mean there is support for these, but not to the extent that they are great for gaming


    -
  • That comment may be insightful, but it is wrong. Windows 2000, in general, performs better at OpenGL benchmarks than Windows 98.
  • No, OpenGL/GLX doesn't give specifications of the card's capabilities (At least last time I looked). Like for example, I might want to enable bump mapping for better quility, but if the card doesn't support that in hardware, I'd prefer to disable it for the speed.

    Let the user decide what he wants enabled and/or disabled, don't dictate. As for capabilities about the card, GL Extensions could be used for this, and would work if card makers created their own GL implimentation, but that rarely happens in the Linux world, It does with with nvidia cards, but for things that need mesa I'm not too sure, anything not hardware accellerated it may do in software.

    Don't know a lot about DRI. Does it still require X?

    It shouldn't require X, the DRM is a kernel interface which anything can use, the dri modules are in the XF86 tree and open, it is theoretically possible that you could write a non-X app to use DRI, but it would be a lot of work I think.

    Strange that the chip makers don't give specs. The whole of the Amiga's custom hardware was public, and there was never a clone of that.

    They're all paranoid, they don't want to lose their cash cow. If another company really wanted to clone your hardware they'd reverse engineer it, register level specs is an insignificant help in that regard.

    -- iCEBaLM
  • by autechre ( 121980 ) on Saturday August 12, 2000 @12:19PM (#860258) Homepage
    The default is for resolutions smaller than the maximum to have a virtual desktop, but this can be turned off, either while running xf86config or editing /etc/X11/XF86Config. Unfortunately, GUI tools such as Xconfigurator (used by RedHat, maybe others) hide this option from you.

    Personally, I think that virtual desktop space is evil, too, especially when I have 3 (or more!) perfectly good other desktops to use (and anyway, the machines I use run at 1024 x 768 or higher).
  • Notice that in the article he mentioned that he compiled all of the drivers from source, and found that the easier route.
    This should not only encourage all those rpm weenies out there to 'use the source, luke, but it also dispells any doubts about the quality of compilation that GCC performs these days.
    I find it quite amusing that an immature open source driver for a free OS, compiled with a freeware compiler can get within a twat filament's distance of the performance of products with literally billions of $$ behind them.
    THAT is a hoot!
    Let's not forget also that GCC is the lingua franca of portable computing. GNU tools are probably the *first* add-ons any sun, hp or aix system receives after setup, so they can actually be *used*.
  • Take a look at the License Agreement (or should it be EULA?) The agreement reads, in part:

    2.1.2 Limitations.

    No Reverse Engineering. Customer may not reverse engineer, decompile, or disassemble the SOFTWARE, nor attempt in any other manner to obtain the source code.

    Now correct me if I'm wrong (IANAL), but aren't there places in which you CAN'T forbid people from reverse engineering and therefore such license terms are void?

    Now what happens if someone (in a perfectly legal fashion) decides to reverse-engineer this driver in one of these countries (in a clean room set-up of course) and publish the source? Will people from non-free countries (USA for instance*) be in legal trouble if they download/use/distribute such new, Free driver?

    * I'll categorize the USA as such until the Supreme Court strikes down the DMCA. If you don't like it... tough. You don't have to read what I write.

  • Comment removed based on user account deletion
  • by sparkmanC ( 93863 ) on Saturday August 12, 2000 @10:24AM (#860274)
    What about the "Time it takes to install" benchmark? New windows drivers are (usually) just a double-click away, while the XFree 4.0 installation is a nasty series of rpms, a kernel patch, and config file editing.

    I hate it when I can't surf slashdot because my X is broken!

    But it's worth it for the little quake icon on my gnome launcher.
  • You have not seen a full version of Win98 for $75 that was legit. Those are OEM cd's are only supposed to be sold with computers.
    Win98 retail is around $189
  • First of all, why the hell was it moderated up to 3? This post is absolutely clueless.

    What you mention (oh, and it's ctrl alt + / ctrl alt - NOT just alt + / alt -) does not change the resolution. The virtual desktop is still stuck at whatever your highest resolution setting in XF86Config is. Well, I guess technically you are right -- the screen resolution does change, but that it not nearly good enough. Even in XFree 4.0 there is NO way to change the resolution AND the virtual desktop size at the same time. There is NO way to change the color depth on the fly either.
    ___
  • Oh well, I guess I learned my lesson by trying to introduce some humor on Slashdot.

    Flamebait? WTF. Overrated I can see, but whoever moderated this as flamebait needs to get a clue.
  • I don't know about you, but I would not be comfortable with a bunch of newbie kids using Linux for games and overlooking its power features. Linux is an OS for the thinking man, not blood-crazed teenagers hungering for first person shooters. Those people should not even have the privilege of using an advanced OS like Linux.

    What exactly do you think the script kiddies are??

  • by doomy ( 7461 ) on Saturday August 12, 2000 @12:03PM (#860288) Homepage Journal
    Well, this is a good start. Perhaps now we can add some of the other things that gamers like - for example an ability to change resolution and depth when not running as root.

    Maybe you were not aware. But I find X to have the simplest and lest obstrusive ways of changing desktop resoultion. If you configured your X correctly have have a number of set resoultions defined as usable on your videocard/monitor combo, then you should be able to change resoltion by just pressting the keys ALT -/+ . Plus to increase, minus to decrease.

    Enjoy
    --
  • by small_dick ( 127697 ) on Saturday August 12, 2000 @12:37PM (#860289)
    Tom's reviews are always hard to read; mostly because of the bad grammer, spelling and logic, (like "....linux is securer than windows 2000...")

    but, he is honest and doesn't hold back. that makes up for it.

    Linux 3d is being handled very poorly. Several years ago, when Linux was struggling through issues with 2D, a lot of people wrote HOWTO's and webpages explaining how to cobble things together. XFree was honest about what did (and did not) work.

    But now, anyone who visits the 3dfx newsgroups or the XFree site, then attempts to get their card working in 3d mode, will notice two things:

    1) There are no HOWTO's or webpages dedicated to your card. When you ask a simple question, you will immediately be bombarded by three morons who say "why didn't you read through the last 712 messages! everything you need is there!"

    2) XFree 4.0.1 documentation claims all sorts of miracles. But the truth is, the product does very little that is new unless you are willing to read the the 712 aforementioned messages and figure out how to disregard the 213 of then that are now outdated or innaccurate.

    Many cards require a CVS download and rebuild of everything, plus a 2.3 series kernel, and many files from other sites. After you build everything and and apply all the hacks, don't be suprised if the your system locks after 20 seconds or so, like mine did.

    Hate to say it, but XFree is really going out on a limb by calling this beast a "3D enabled release". It's a immature, undocumented 3D release being supported by a bunch of monkeys who refuse to properly document the 3D workarounds. The later 3.X releases were better documented, both by XFree and the card vendors, and more reliable in 3D mode.

    The 2D side of 4.0.1 is stable, and the new server configuration stuff is wonderful. But this is not a quality 3D release by any means.
  • by 91degrees ( 207121 ) on Saturday August 12, 2000 @10:31AM (#860294) Journal
    Absolutely. Text based games are the way to go. In fact, I recenty spent $179.99 on a BSD Star Trek accelerator. Quite an investment, but you should see how fast it goes. Especially if I reduce the res to 40 columns. And Hangman is incredible. Even some of the serious apps in /usr/games/bin like morse have increased their speed considerably.
  • by Moridineas ( 213502 ) on Saturday August 12, 2000 @10:32AM (#860296) Journal
    I would like to see benchmarks of the DRI drivers that are more crossplatform (work being done on FreeBSD for example) as well as open and supported by vendors like ATI,Matrox, and 3dfx.
  • by ywwg ( 20925 ) on Saturday August 12, 2000 @10:33AM (#860297) Homepage
    I don't care how many articles are written about how great gaming on linux is, it is still _unplayable_ without decent mouse support. I have the latest drivers from nvidia that work with xfree 4.0.1. I've tried everything from Option SampleRate to tuneps2.

    In short, NOTHING WORKS. the response in Q3A is slow and jerky, unless you are standing stock still. If you turn m_filter off in windows, the mouse is still smooth because the ps/2 rate acn be cranked so high. In linux with m_filter off, it is sucky sucky sucky.

    Somebody has to get Loki and XFree in a room together and FIX THIS. Check the newsgroups. check slashdot. _everyone_ complains about this, but nothing gets done. I'm tired loki saying "oh, we support DGA, it should work," and XFree saying, "we fixed it in 4.0.1." Can somebody _please_ address this problem, once and for all?

    </rant> sorry.
  • by account_deleted ( 4530225 ) on Saturday August 12, 2000 @10:35AM (#860298)
    Comment removed based on user account deletion
  • Well, since XFree 4.0 isnt part of the standard distributions yet and even when it is its going to be a 'first version', thats what you get.

    When you do get support tho; well, lets just say that it took me about 10 minutes to install a Voodoo 3 PCI card under linux (plug in, boot, autodetect), and three _days_ to get the same card working under Windows '98 (some bug, driver problem, downloads, OS reinstalls, conflicts with old mobo-mounted gfx chipset and drivers). It still locks up with a black screen and goes temporarily unpingable for two minutes every time I boot windows.

    YMMV, of course, but in my experience if you check that you arent buying unsupported hardware it's usually easier these days to get it running under linux than it is under windows ('course, sometimes Windows 'Just Does What I Tell It To' too).
  • by be-fan ( 61476 ) on Saturday August 12, 2000 @01:21PM (#860303)
    Fix the bloody kernel and volia, the problem is gone. It's the truth. All hardware vendors give up working on drivers eventually. Even OSS hardware drivers will not get updated when they get really old, simply because there are bigger fish to fry. Having a stable driver API is a big help here. Really, though, NVIDIA is really good about old hardware. I have an original TNT and they're still giving upgraded drivers for it.
  • There are some problems with specific cards (particulary NVIDIA cards.) However, the NT series really DOES perform faster. Even for simple programs like mine NT 4.0 is usually 10-20% faster than Win9x.
  • This would be a troll, if this guy were lying. Unfortunately for the Linux crowd, he's right. NT4 doesn't get 30% better fps, but it is significantly faster than Win9x.
  • Uh, the binaries that come in packages such as RPM are most often compiled using GCC. This means that installing from either source or RPM gets you about the same quality of a compilation and performance. The part that makes RPMs undesirable at times is the variety of directory trees and configurations out and about. A package designed for RedHat might not work in SuSE, Caldera, or your home spun distro. Go hoot yourself.
  • by GauteL ( 29207 ) on Sunday August 13, 2000 @12:55AM (#860312)
    The reason 60fps is better than 30fps
    is that we're talking average here.
    With 30fps the minimum could end up being
    18 fps (which is noticable), while with
    60 fps the minimum might be around 40 fps.
    If you had a game which held 30fps AT ALL TIMES,
    it would be just as good as 60fps ALL THE TIME.
  • by be-fan ( 61476 ) on Saturday August 12, 2000 @01:32PM (#860313)
    The NVIDIA Linux driver is faster than it looks from some of the benchmarks. The thing he mentioned, the page-flipping vs. blitting issue, seems to account for most of the performance difference between linux and Win98. I say this, because under the low-res Solar System tests, Win98 and Linux perform almost equally. Tom attributes this to CPU limitations, but the GeForce cards are geometry accelerators, so CPU limitations really shouldn't be an issue here. If the NVIDIA Linux driver really were slower, than transforms would be slower as well, and thus the Solar System score would be lower. Same thing for the low res Quake III tests.

    That brings me to a question. Why doesn't the NVIDIA Linux driver implement page flipping. Page-flipping is a basic necessity (nay, a innate right!) for game developers. I seriously doubt X doesn't allow access to page flipping in full screen mode... does it? Also, I just thought of something. Does X allow access to page-flipped overlays. That might allow Quake to use page-flipping in window'ed modes. (Which would be pretty cool.)
  • by be-fan ( 61476 ) on Saturday August 12, 2000 @01:41PM (#860315)
    As far as I recall, Windows 98 has protected memory, preemptive multi-tasking, and the whole OS is basically a user-land library. (Actually, by making the OS mainly consist of a set of DLL's loaded by the application, performance is IMPROVED, since API calls don't have to do a switch into Ring0. It also means that Win98 is in a way a pseudo-exo-kernel OS.)
    NT actually performs better in Quake than Win98 does.
    Windows2000 solves the whole "big OS/gaming" problem quite nicely. You see, the OS needs of games are quite limited. Thus, in effect, DirectX is really as much of an OS as most games need. Since you are rarely being productive when you're gaming, some modes of DirectX allows a program to hog the system at the expense of other applications. However, since you're not using other applications while you're gaming, this isn't a problem. Quite an elegant solution really. Although, I'm against heavy OSs in general, but within the context of the problem, Win2K's solution is a pretty good one. (Though don't get me started on the code bloat. What I want is NT4 with full DirectX. Is that too much to ask?)
  • by MostlyHarmless ( 75501 ) <artdent@[ ]eshell.org ['fre' in gap]> on Saturday August 12, 2000 @10:44AM (#860320)
    Linux didn't used to be about money ... While many people in the Linux community, including myself, didn't welcome RedHat's IPO, it shows how valuable Linux has become

    Why, oh why, do people always think that open-source or software liber implies that nobody can make money off of it? Contrary to many peoples' opinions, RMS does not mind if you make money. Why is it bad that RedHat went IPO? What can they possibly do to the community? The very nature of the GPL is that no one person can run off with the code. There is no chance that they would make the kernel non-free or anything like that, like many people were claiming would happen. Instead, what has happened to the free software community as a result of the IPO?

    They've hired programmers to work on the kernel
    They've hired programmers to work on GNOME
    They've helped bring the ideas of the open-source community to the masses.

    The last is the most important of all. Hiring programmers has been a great help to ensuring the success of critical parts of the GNU/Linux system, but that by itself would have meant very little without their support of the free software community. Without freedom, it would have just been another software project. Instead, their help of the GNOME project as well as the kernel has helped prepare the platform for their most important aid to our community. That aid was in spreading our ideas to the world. Yes, they are taking more of an open-source pragmatist approach as opposed to the free software idealistic approach. But even exposure to the former will help, and eventually some of the users will open their minds towards the ethics of free software, not just the business of open-source.

    Sorry if this turned out to be a rant.
    --

One man's constant is another man's variable. -- A.J. Perlis

Working...