Slashdot Mirror


Major Performance Improvement Discovered For Intel's GPU Linux Driver

An anonymous reader writes: LunarG, on contract with Valve Software, discovered a critical shortcoming with the open-source Intel Linux graphics driver that was handicapping the performance. A special bit wasn't being set by the Linux driver but was by the Windows driver, which when enabled is increasing the Linux performance in many games by now ~20%+, which should allow for a much more competitive showing between Intel OpenGL performance on Windows vs. Linux. However, the patch setting this bit isn't public yet as apparently it's breaking video acceleration in certain cases.

96 comments

  1. Will it work with my distro? by NotDrWho · · Score: 4, Funny

    Because my distro is the best.

    --
    SJW's don't eliminate discrimination. They just expropriate it for themselves.
    1. Re:Will it work with my distro? by Anonymous Coward · · Score: 1

      Fuck you, mine is better.

    2. Re:Will it work with my distro? by Chrisq · · Score: 4, Funny

      Fuck you, mine is better.

      I'm waiting for it to be ported to HURD

    3. Re:Will it work with my distro? by i+kan+reed · · Score: 1

      It will if you go and set the bit yourself, and correct all the errors it causes.

    4. Re:Will it work with my distro? by Pope+Hagbard · · Score: 3, Funny

      Xemacs is the best Linux distribution.

    5. Re:Will it work with my distro? by Anonymous Coward · · Score: 5, Funny

      I agree. The only thing it is missing is a good text editor.

    6. Re:Will it work with my distro? by Anonymous Coward · · Score: 5, Funny

      As a longtime user of both "My Distro" Linux and "Mine" Linux, I can honestly say that I have never noticed a performance difference between them. However, I can emphatically state that "Your Distro" Linux sucks.

    7. Re:Will it work with my distro? by Immerman · · Score: 1

      Better get comfortable, maybe have a goblet of wine. Better yet bring a few barrels.

      --
      --- Most topics have many sides worth arguing, allow me to take one opposite you.
    8. Re:Will it work with my distro? by NotDrWho · · Score: 2

      I bet you're not even using the best Desktop, which is the one I use.

      --
      SJW's don't eliminate discrimination. They just expropriate it for themselves.
    9. Re:Will it work with my distro? by Anonymous Coward · · Score: 0

      It will, if you install the correct text editor and uninstall that stupid one.

    10. Re:Will it work with my distro? by Wootery · · Score: 3, Informative

      Wine? Valve are supporting Linux natively!

    11. Re: Will it work with my distro? by Anonymous Coward · · Score: 0

      Alright, that's done it.

      Now, someone has to create My Distro.

    12. Re:Will it work with my distro? by occasional_dabbler · · Score: 0

      Oh, for mod points today! Well played, sir, you made my evening :-)

      --
      "Our opponent is an alien starship packed with atomic bombs," I said. "we have a protractor"
    13. Re:Will it work with my distro? by Sun · · Score: 1
    14. Re:Will it work with my distro? by Wootery · · Score: 2

      Huh. 4 Informative? I was going for Funny.

    15. Re: Will it work with my distro? by toddestan · · Score: 1

      Will it install on My Computer?

    16. Re:Will it work with my distro? by Anonymous Coward · · Score: 0

      That's what evil mode is for.

  2. Benchmark Bit by bluefoxlucid · · Score: 5, Funny

    It's the Benchmark Bit!

    1. Re:Benchmark Bit by TheRealMindChild · · Score: 5, Insightful

      Probably the "Official driver" bit

      --

      "When life gives you lemons, don't make lemonade. Make life take the lemons back!" -- Cave Johnson
    2. Re:Benchmark Bit by Anonymous Coward · · Score: 5, Informative

      The Intel drivers for Linux are official and open source. They are actively maintained by Intel themselves. This is not like the Nvidia/Nouveau split, Intel are actually very open source friendly in this area.

    3. Re:Benchmark Bit by UnknowingFool · · Score: 3, Interesting

      Also most likely the Linux drivers are written by different people than the Windows or OS X drivers so there are probably so noted differences in the coding. Some of the differences are due to APIs. Procedurally there would be practical reasons to keep driver development separate. There may be some NDA stuff on the proprietary side that Intel has to maintain with the OS.

      --
      Well, there's spam egg sausage and spam, that's not got much spam in it.
    4. Re:Benchmark Bit by arglebargle_xiv · · Score: 1

      The Intel drivers for Linux are official and open source. They are actively maintained by Intel themselves. This is not like the Nvidia/Nouveau split, Intel are actually very open source friendly in this area.

      So you've got the choice between crappy graphics hardware with OSS drivers and high-end graphics hardware with binary-blob drivers. Damn.

    5. Re:Benchmark Bit by Anonymous Coward · · Score: 0

      Intel are actually very open source friendly in this area.

      There STILL is no usable Linux graphics driver for Intel systems built with PowerVR graphics ...

  3. The "paid Microsoft tax" bit, apparently by pla · · Score: 4, Insightful

    I notice TFA has almost no detail beyond what TFS says. Yeah, so they found this bit that apparently has no side effects to anything else but magically boosts performance by 20%? I'll admit I haven't written a graphics card driver since back in the VESA2 days, but I can't even conceive of what function such a bit could have, without having some down side... Something like (and I don't mean this literally) disabling vsync but accepting tearing.

    1. Re:The "paid Microsoft tax" bit, apparently by Anonymous Coward · · Score: 5, Informative

      You read the article, but did you read the summary?

      "However, the patch setting this bit isn't public yet as apparently it's breaking video acceleration in certain cases."

    2. Re:The "paid Microsoft tax" bit, apparently by Anonymous Coward · · Score: 0

      Most likely it's the, "trust me, I know better than you, you can do this operation without overheating" bit. And if you set it on all the time, the chip overheats, hence braking functionality in some cases.

    3. Re:The "paid Microsoft tax" bit, apparently by jones_supa · · Score: 2

      On Haswell architecture there is some special bit that has to be set for the hardware samplers to work at maximum performance. There seems to be no details regarding what this specifically means. The original LunarG post seems to contain the best information. What is quite cool is that they claim to have plugged LLVM into the Mesa shader compiler.

    4. Re:The "paid Microsoft tax" bit, apparently by jeffmeden · · Score: 2

      I notice TFA has almost no detail beyond what TFS says. Yeah, so they found this bit that apparently has no side effects to anything else but magically boosts performance by 20%? I'll admit I haven't written a graphics card driver since back in the VESA2 days, but I can't even conceive of what function such a bit could have, without having some down side... Something like (and I don't mean this literally) disabling vsync but accepting tearing.

      It disables DirectX 7 backward compatibility......

      think i am kidding?

    5. Re:The "paid Microsoft tax" bit, apparently by Impy+the+Impiuos+Imp · · Score: 1

      Which is irrelevant if its purpose is to look for and accelerate benchmark-like activities.

      Is this perma-set on Windows? Or just on benchmarks? And what would the bit be called?

      --
      (-1: Post disagrees with my already-settled worldview) is not a valid mod option.
    6. Re:The "paid Microsoft tax" bit, apparently by Anonymous Coward · · Score: 0

      Maybe it's something DTACK GROUNDED on the 68K. Great performance boost if your hardware can support it.

      There are many ways to "cheat" when drawing (think of Bressenham lines). The cheats may not be visible in some things (such as a game) but easily detected in others. Even something as simple as enabling/disabling a codec can have downstream impact.

    7. Re:The "paid Microsoft tax" bit, apparently by mrzaph0d · · Score: 2

      it's "turbo_bttn"

      --
      this is just a placeholder till i send back my real sig from the future.
    8. Re:The "paid Microsoft tax" bit, apparently by TheRaven64 · · Score: 2

      I don't know specifically, but for some GPUs there's a flag to disable strict IEEE floating point compliance. This can make things go a lot faster, at the expense of sometimes giving the wrong result. For graphics workloads, a few more floating point rounding errors are normally invisible to the human eye, but for scientific (GPGPU) computing they can be problematic.

      --
      I am TheRaven on Soylent News
    9. Re:The "paid Microsoft tax" bit, apparently by Anne+Thwacks · · Score: 1

      Its the "Windows is running" bit.

      --
      Sent from my ASR33 using ASCII
    10. Re:The "paid Microsoft tax" bit, apparently by Compholio · · Score: 1

      ... conceive of what function such a bit could have, without having some down side..

      Power-save -> 0, Performance -> 1

    11. Re:The "paid Microsoft tax" bit, apparently by pla · · Score: 1

      Power-save -> 0, Performance -> 1

      Laptop-safe -> 0, crotch-burn -> 1. :)

      But yeah, I see your point, that it wouldn't have any side effects in the output-integrity sense.

    12. Re:The "paid Microsoft tax" bit, apparently by Anonymous Coward · · Score: 0

      Turns off the DRM?

    13. Re:The "paid Microsoft tax" bit, apparently by Anonymous Coward · · Score: 0

      hence *braking* functionality in some cases.

      You mean it slows it down again?

    14. Re:The "paid Microsoft tax" bit, apparently by Anonymous Coward · · Score: 0

      It's a "chicken bit":

      http://stackoverflow.com/questions/326880/what-are-the-best-practices-for-hardware-description-languages-verilog-vhdl-et

      [i]Have "chicken bits". Chicken bits are bits in MMIO set via the driver to disable a feature in silicon. It's intended to revert changes made in which confidence is not high (confidence is directly proportional to validation efforts). It is next to impossible to hit every possible state in pre-silicon. Confidence on your design cannot truly be met until it's proven in post-silicon. Even if there is only 1 state that is hit 0.000005% of the time that exposes the bug, it WILL HIT in post-silicon, but not necessarily in pre-silicon.[/i]

      Looks like this bit does a wait-for-sync with the video decompressor, framebuffer blitting and flushing. Most times, an entire frame will be decompressed on time, but there are odd occasions where it takes longer than expected. Without this bit set, the video would tear. With this bit set, the video would render correct, but other rendering operations would take a performance hit as they went through the
      sync.

    15. Re:The "paid Microsoft tax" bit, apparently by Anonymous Coward · · Score: 0

      I notice TFA has almost no detail beyond what TFS says. Yeah, so they found this bit that apparently has no side effects to anything else but magically boosts performance by 20%? I'll admit I haven't written a graphics card driver since back in the VESA2 days, but I can't even conceive of what function such a bit could have, without having some down side... Something like (and I don't mean this literally) disabling vsync but accepting tearing.

      It disables DirectX 7 backward compatibility......

      think i am kidding?

      Um . No. I don't.

    16. Re:The "paid Microsoft tax" bit, apparently by Anonymous Coward · · Score: 0

      Same A.C. again.

      Is it true that the principle use of A.C. is to hide your identify from your employer rather than from Slashdotters?

  4. now you've done it... by funkymonkjay · · Score: 5, Funny

    you've flipped the EVIL bit!

    1. Re:now you've done it... by NotDrWho · · Score: 2

      Yeah, there's your problem right there. Someone set this driver to "evil." ;-)

      --
      SJW's don't eliminate discrimination. They just expropriate it for themselves.
    2. Re:now you've done it... by funkymonkjay · · Score: 1

      Evil Inside (tm)

    3. Re:now you've done it... by Anonymous Coward · · Score: 0

      No no no. It's the Dirty bit!

  5. "apparently it's breaking video acceleration in ce by grahamtriggs · · Score: 5, Insightful

    That's probably why it wasn't being set then.

  6. All games? by Anonymous Coward · · Score: 2, Funny

    Would this help with nethack?

    1. Re:All games? by Anonymous Coward · · Score: 1

      Yes, but the grid bug kills you 20% faster.

    2. Re:All games? by Anonymous Coward · · Score: 0

      My own fastest nethack game was: move, move, receive a rock on the head. dead.

  7. The Turbo Button by CambodiaSam · · Score: 2

    Someone found the Turbo Button from my old 386. http://en.wikipedia.org/wiki/T...

    1. Re:The Turbo Button by Carewolf · · Score: 1

      Someone found the Turbo Button from my old 386. http://en.wikipedia.org/wiki/T...

      Funny thing with the turbo button was that it made your computer slower. Specifically as slow as a 8086.

    2. Re:The Turbo Button by Anonymous Coward · · Score: 0

      Yes, but a consumer is going to buy the PC with a turbo button, not a slow button.

    3. Re:The Turbo Button by Anonymous Coward · · Score: 0

      They could just as easily made the 'OFF' position the compatibility (eg: slow speed) position, and the 'ON' position the fast (eg: full speed) position.

    4. Re:The Turbo Button by Anonymous Coward · · Score: 0

      All I remember was a display that said 100MHz or 133MHz. You pressed the button whichever way made it 133MHz. Were they lying to me?!

    5. Re:The Turbo Button by Nyder · · Score: 1

      Actually it depended on the motherboard what the "Turbo" button did. Some it was for speed ups, some it was for speed downs, and some it didn't even work.

      --
      Be seeing you...
    6. Re:The Turbo Button by Anonymous Coward · · Score: 0

      Hey look, Car 1 is going faster than car 2! No way, man. Car 2 is going slower than Car 1!

    7. Re:The Turbo Button by canadiannomad · · Score: 1

      and some it didn't even work.

      Probably because people like me wouldn't bother connecting the button on computers we built, and would instead just put a jumper.

      --
      Hmm, the humour and sarcasm seem to have been be lost on you.
    8. Re:The Turbo Button by toddestan · · Score: 1

      That's exactly how it worked on every computer I ever came across. For normal operation you'd always have the computer in 'turbo'.

  8. This one simple bit bumps your performance by 20% by Overzeetop · · Score: 5, Funny

    ...and you won't believe how easy it is!

    C'mon editors, it's like you missed the entire social media headline writing class.

    --
    Is it just my observation, or are there way too many stupid people in the world?
  9. I have seen that before by Anonymous Coward · · Score: 3, Insightful

    Removing flush() from code boosts DBMS performance by 40% but that patch wasn't yet accepted as in rare cases it makes data disappear after power failures.

  10. ...in certain cases. by Thanshin · · Score: 2

    So, it works 20% better... until it crashes.

    And... we all know most of the Linux user base would want to give up random crashes for 20% better performance, right?

    1. Re:...in certain cases. by Anonymous Coward · · Score: 0

      No, but Windows users would never notice more frequent crashing ;)

    2. Re:...in certain cases. by Anonymous Coward · · Score: 0

      That's the idea behind systemd, innit?

    3. Re: ...in certain cases. by Anonymous Coward · · Score: 0

      STFU

  11. Re:Pointless improvement? by Anonymous Coward · · Score: 0

    I use intel graphics on my laptop for gaming.
    The 4400 is pretty good for playing something while waiting around at the airport or in the hotel.
    For real gaming I have my PC at home.

  12. Re: Pointless improvement? by bill_mcgonigle · · Score: 4, Insightful

    People are quick to yell "games" as if the entire desktop hasn't been 3D-accelerated for a decade.

    In many cases if a chip can be done 30% faster, not only is the user happy with the visuals, but that silicon can enter low-power mode more quickly. A laptop user might get a few more minutes' battery life with this bit on and the world might burn a few less tons of coal for all of the systems.

    --
    My God, it's Full of Source!
    OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
  13. What was the summary in English? by Anonymous Coward · · Score: 0

    It sure doesn't scan well...

  14. Re:"apparently it's breaking video acceleration in by Anonymous Coward · · Score: 2, Insightful

    Your comment is misleading. Just because a performance boost causes stability issues (whilst still under development!) doesn't mean those issues can't be ironed out. If the bit works on Windows, most likely it will work on Linux too after the devs make the right adjustments. The fact that it wasn't set was a mistake, not a design decision.

  15. Re: Pointless improvement? by Anonymous Coward · · Score: 0

    People are quick to yell "games" as if the entire desktop hasn't been 3D-accelerated for a decade.

    Well of course, this is Linux we're talking about. They're not going to pass up an opportunity to pretend as if one of Linux's major shortcomings doesn't actually exist.

  16. Re: This one simple bit bumps your performance by by Anonymous Coward · · Score: 1

    One weird bit, benchmarks hate it

    Yeah, yeah keep reaching for that low hanging fruit, beating that dead horse.

  17. Excuse me? by Anonymous Coward · · Score: 0

    It's the best GNU distribution.

  18. Re:Pointless improvement? by Anonymous Coward · · Score: 0

    Real Gamers have nothing (positive) to say, and let everyone know it.

    Real Gamers waste Real Money on Fake Crap.

    Real Gamers would not be Really Missed if their entire subculture of masturbatory fantasies vanished tomorrow.

    Leigh Alexander is that you ?

  19. Re: This one simple bit bumps your performance by by Andrewkov · · Score: 1

    The secret bit Intel doesn't want you to know about!

  20. Whatever by future+assassin · · Score: 1

    with this new bit switch my BeOS can now perform multimedia tasks with the power of a uber computer.

    --
    by TheSpoom (715771) Uncaring Linux user here. I have nothing to add to this but please continue. *munches popcorn*
  21. Re: Pointless improvement? by jedidiah · · Score: 2

    > People are quick to yell "games" as if the entire desktop hasn't been 3D-accelerated for a decade.

    Except you don't need a good gaming card in order to deal with that crap. The cheapest trailing edge AMD/nvidia stuff that's likely to be not supported in the next driver release is more than fast enough for "desktop 3D acceleration".

    --
    A Pirate and a Puritan look the same on a balance sheet.
  22. A question: would this affect CUDA performance? by whitroth · · Score: 1

    Some of us are using those GPUs for actual work - like serious scientific computation. Anyone know anything about how this bit switch would affect CUDA?

                    mark

    1. Re:A question: would this affect CUDA performance? by Anonymous Coward · · Score: 1

      Some of us are using those GPUs for actual work - like serious scientific computation. Anyone know anything about how this bit switch would affect CUDA?

                      mark

      CUDA on intel processors? "Science guy", you don't know what you're talking about

    2. Re:A question: would this affect CUDA performance? by Anonymous Coward · · Score: 0

      You will get your answers twice as fast, but it will be incorrect half the time. Enjoy the future!

    3. Re:A question: would this affect CUDA performance? by Anonymous Coward · · Score: 0

      CUDA doesn't work on Intel GPUs. It's an nVidia thing. Which you'd know if you ever tried to use it.

    4. Re:A question: would this affect CUDA performance? by Anonymous Coward · · Score: 0

      I'm pretty certain that CUDA is an NVIDIA thing, not an Intel thing. So, umm, its not going to do anything for Nvidia hardware, since this bit is for Intel hardware. The thing you have to understand is: not all hardware is exactly the same. The engine in a Ford Tempo isn't the same engine as a Ferrari Testarossa, and so setting a 'bit' in the Testerosa might have meaning, but attempting to do the same thing in the Tempo may have no function at all (there may not even be an address there, there might be nothing physical there either). Its like having an enable bit for the 11th and 12th cylinders, and the Tempo is only has a 4 cylinder engine. So you want to apply what is going on with the Intel graphics card somehow to an Nvidia graphics card..... Are you sober? Were you ok to drive when you posted the question? Its Friday, Happy Hour rules apply: no posting if you need a pole to stay upright.

    5. Re:A question: would this affect CUDA performance? by godrik · · Score: 1

      Since when does Intel ship devices that support nVidia CUDA?

  23. Re:Pointless improvement? by Guspaz · · Score: 2

    Real gamers often do their gaming on a desktop, and have an Ultrabook for portability. Said gamers might not want or need the bulk of a gaming notebook 99% of the time, but might still appreciate the ability of an Intel iGPU to handle basic game rendering on the rare occasion when they want to keep themselves busy while on the go.

    I'm a gamer, and I do all my gaming on a relatively high-end desktop. I've got a Macbook Air, because I only have a desire to fire up a game on my notebook a handful of times a year. But at the same time, I appreciate that I can run Civ V or Civ: BE on said notebook when it's called to do so.

  24. I liked the original headline better by Minwee · · Score: 3, Funny

    "How A Mom Found A Secret Bit That Improves Her Video Driver Performance By Twenty Percent! Intel Hates Her!"

    1. Re:I liked the original headline better by Spamalope · · Score: 1

      "How A Mom Found A Secret Bit That Improves Her Video Driver Performance By Twenty Percent! Intel Hates Her!"

      Using one weird trick? I'll never believe what happened next!

  25. Re:This one simple bit bumps your performance by 2 by mccrew · · Score: 2

    I flipped this bit, and you won't believe what happened next!

    --
    Hey, Windows users, there is no such thing as "forward" slash, there is only slash and backslash.
  26. Re: This one simple bit bumps your performance by by Overzeetop · · Score: 1

    Oooh, that's a good one.

    Man, I hate these headlines. I now reflexively ignore stories which are captioned this way - basically anything which doesn't describe the article content, no matter how many boobs, muscles, butts, or cute/sad animals are in the photo caption. No, I take that back - *especially* if there are boobs, muscles, butts, or cute/sad animals are in the photo caption.

    --
    Is it just my observation, or are there way too many stupid people in the world?
  27. Re: "apparently it's breaking video acceleration i by SpaghettiPattern · · Score: 1

    But _is_ set in the Windows driver? No sense in that.

    --

    I hadn't the slightest objection to his spending his time planning massacres for the bourgeoisie... (P.G. Wodehouse)
  28. Re: Pointless improvement? by Anonymous Coward · · Score: 0

    I beg to differ. Low-end nVidias can't handle 1080p composited desktops. The amount of bandwidth being burned by storing all the windows' contents into textures is considerable, and they just lack the bandwidth necessary for smooth operation. Just try an oldish GF 8200.

  29. Re:This one simple bit bumps your performance by 2 by Anonymous Coward · · Score: 1

    Windows developers HATE this man!

    Developer discovers this one "weird" trick to boost driver performance.

  30. dirty bit? by Anonymous Coward · · Score: 0

    dirty bit?

  31. Re: Pointless improvement? by Anonymous Coward · · Score: 0

    GeForce 8000 cards are nearing a decade old at this point. I'd be surprised if today's version of "Intel integrated graphics" didn't surpass it in the relevant benchmarks (Scoring high FPS in a high-polycount scene is not a relevant benchmark)

  32. Re: "apparently it's breaking video acceleration i by m.alessandrini · · Score: 1

    Because on windows people don't complain when the system crashes.

  33. Re: "...waiting for it to be ported to HURD" by peacefool · · Score: 1

    It's hurdly possible. Sorry.