Forgot your password?
typodupeerror
AMD Graphics Portables Technology

AMD Launches World's First Mobile DirectX 11 GPUs 169

Posted by kdawson
from the forty-nm-of-fast dept.
J. Dzhugashvili writes "Less than 4 months after releasing the first DX11 desktop graphics card, AMD has followed up with a whole lineup of mobile graphics processors based on the same architecture. The new Mobility Radeon HD 5000 lineup includes four different series of GPUs designed to serve everything from high-end gaming notebooks to mainstream thin-and-light systems. AMD has based these processors on the same silicon chips as its desktop Radeon HD 5000-series graphics cards, so performance shouldn't disappoint. The company also intends to follow Nvidia's lead by offering notebook graphics drivers directly from its website, as opposed to relying on laptop vendors to provide updates."
This discussion has been archived. No new comments can be posted.

AMD Launches World's First Mobile DirectX 11 GPUs

Comments Filter:
  • by Anonymous Coward

    Who the hell other than the poor sods still doing x86 Windows only game/graphics development still uses that turd of an API DirectX?

    Let's just go over the platforms I work on:

    PC graphics development - OpenGL
    Linux graphics development - OpenGL
    Mac graphics development - OpenGL
    Android graphics development - OpenGL ES
    iPhone graphics development - OpenGL ES
    Embedded ARM based system development - OpenGL ES

    even some OpenGL for console development.

    • Re: (Score:3, Informative)

      by confused one (671304)
      And that makes perfect sense if you're targeting all those different platforms. There may even be perfectly reasonable reasons to use OpenGL over DirectX based on your coding requirements and the APIs. However, if you're target audience is Window and Windows Embedded only, and there are no requirements that are better served by OpenGL, there's no reason not to use DirectX.

      It's just a tool.
      • by Lodragandraoidh (639696) on Thursday January 07, 2010 @06:59PM (#30688888) Journal

        I read your post and it occurred to me that it illustrates perfectly a key problem with software development today: short sightedness.

        In an age of fast multiprocessing, it only makes sense to do everything you can to create abstraction layers that will ensure:

        1. My software will have the widest possible audience regardless of platform. $$$

        2. I will be able to extend the application, or create a new one with minimal effort by reusing modules I've already created to do hard things well/fast. $$$ (in form of turn-around time/effort)

        3. If a vendor decides to break something in their firmware/hardware - I only have to fix one module that drives the given hardware - *NOT* the application itself. $$$ (ditto)

        Flexibility, resiliency, more cash in your pocket...I don't see a down side to taking this approach. On modern gaming rigs in particular, there is no reason NOT to use OpenGL - for all it's perceived limitations compared to a tweaked out directX X86 app.

        As a gamer myself, I look at it from another angle: I have Linux, Mac machines as well as a high-end Windows game rig - to host games (I like to create and share my own maps/scenarios in some games) cost efficiently I prefer to use the Linux server, and play on my Windows box....using and tweaking WINE in order to run the game (I'm not made of money and can't cost-justify a full compliment of windows servers - which also would waste resources since I am a *nix developer too). Getting WINE to work with some of the niche games I play is a royal pain. If the developers of said games took my advice, I would be running their games natively under linux with minimal headaches.

        Flexibility and choice is good for the widest audience. Vendor lock-in is bad - and only serves a few types of people (the corporation$$$ and simple gamer-$$$). The funny thing is, these companies stand to make more money than they would under their lock-n strategy if they would think long term and build flexible extensible applications that benefit the largest audience. Lucky for me most of the titles I currently enjoy have taken this approach; I will continue to gravitate to those that do, and deny $$$ to those that won't.

        • by Chryana (708485)

          This all sounds good in theory, but I don't think it works in practice.
          - The Linux user is really small, so there's not much point in developing games for them; and worse, we're all tech-savvy, so most Linux gamers have a Windows computer somewhere if there is a game they want to run. So your first point is completely moot.
          - I don't think game development benefits nearly as much from code reuse as most other software development.

          Anyways, I think the current state of game development speaks in my favor. Take

        • by RAMMS+EIN (578166)

          The question is, with all your arguments favoring OpenGL over Direct3D, why are developers still using the latter?

          And why do the developers of many portable 3D libraries and game development libraries (Crystal Space, OGRE, etc.) support Direct3D, even though they already have OpenGL support, which would supposedly work just fine?

        • You ignore that it also takes more time and resources.
          There's a marketing window for most projects, whether it's a tie-in to a sport/movie/show or making sure your game isn't considered outdated upon release.
          Software as a business is about releasing a product that is "good enough." Spending more time costs money, both in actual cost and potential costs (you could have the same dev team working on another project).
          Tweaking may get more eyes, but it doesn't necessarily mean more money.
    • Re: (Score:3, Insightful)

      by MBGMorden (803437)

      Yeah those "poor sods" making multi-million dollar grossing titles. Seriously, I'm all for OpenGL. I like it because it does make ports easier and I'd like to see more games available on Linux and Mac.

      The snide "are people STILL using technology X?" comments when technology X is the clear market leader are just annoying though.

    • by Sycraft-fu (314770) on Thursday January 07, 2010 @05:29PM (#30687810)

      As well as a good deal of other Windows graphic programs. You can stick your head in the sand and pretend that Microsoft Windows isn't a major player, but you are fooling only yourself. Windows development matters a whole lot, and DX is the native API and thus many use it.

      However, in this case the reference is to features of the card. See OpenGL is really bad about staying up to date with hardware. They are always playing catchup and often their "support" is just to have the vendors implement their own extensions. So when a new card comes out, talking about it in terms of OpenGL features isn't useful.

      Well, new versions of DirectX neatly map to new hardware features. Reason is MS works with the card vendors. They tell the vendors what they'd like to see, the vendors tell them what they are working on for their next gen chips and so on. So a "DX11" card means "A card that supports the full DirectX 11 feature set." This implies many things, like 64-bit FP support, support for new shader models, and so on. IT can be conveniently summed up as DX11. This sets it apart to a DX10 card like the 8800. While that can run with DX11 APIs, it doesn't support the features. Calling it DX10 means it supports the full DX10 feature set.

      So that's the reason. If you want to yell and scream how OpenGL should rule the world, you can go right ahead, however the simple fact of the matter is DirectX is a major, major player in the graphics market.

      • by RAMMS+EIN (578166)

        ``However, in this case the reference is to features of the card. See OpenGL is really bad about staying up to date with hardware.''

        How can that be, when it allows vendors to add their own extensions? Add a feature to your hardware, add an extension to OpenGL so programmers can use it. No need for delays.

        ``They are always playing catchup and often their "support" is just to have the vendors implement their own extensions.''

        Is there a problem with that? I mean, yes, it would be nicer if features were immedia

        • Re: (Score:3, Informative)

          by Arterion (941661)

          Well... it IS easier to buy a video card that says DX10, and know that a game that says DX10 is going to run on it. Trying to keep up with all the extensions your card is going to support or not when you're at the store looking at games on a shelf would be a nightmare.

    • Xbox 360 graphics development - DirectX
      XNA (Xbox 360 indie games) graphics development - a managed API based on DirectX

    • by Kjella (173770) on Thursday January 07, 2010 @06:02PM (#30688232) Homepage

      Only all the AAA games on Windows, but clearly you are far more important than them.

    • Re: (Score:3, Informative)

      Who the hell other than the poor sods still doing x86 Windows only game/graphics development still uses that turd of an API DirectX?

      I know you've specifically excluded Carmack here, but nonetheless, I think his opinion is not exactly irrelevant:

      "DX9 is really quite a good API level. Even with the D3D side of things, where I know I have a long history of people thinking I'm antagonistic against it. Microsoft has done a very, very good job of sensibly evolving it at each step—they're not worried about breaking backwards compatibility—and it's a pretty clean API. I especially like the work I'm doing on the 360, and it's probabl

  • by NotBorg (829820) on Thursday January 07, 2010 @05:04PM (#30687500)

    DirectX 11 in a mobile device? So the device doubles as a hairdryer?

    • Re: (Score:3, Informative)

      by mikael (484)

      Embedded systems may only be using a screen resolution of 640x480 or 800x600 rather than dual monitor 2048x1536. That's one energy/time saving. Then there won't be 900+ stream processors like the high-end gaming cards, there might just be 128 or 256. There's another saving. Anti-aliasing will be disabled as well, so that saves some processing time and power as well.

      You will still have texture mapping, shadowing effects using fragment shaders, but just not as many triangles as the current gaming engines will

  • by MostAwesomeDude (980382) on Thursday January 07, 2010 @05:10PM (#30687572) Homepage

    Support in the open-source drivers is being written as fast as ATI can verify and declassify docs. Also the r600/r700 3D code should be mostly reusable for these GPUs.

    • Re: (Score:3, Interesting)

      by owlstead (636356)

      How many years was it again that they promised to produce open source graphic drivers for Linux? I've lost count and have ordered a new motherboard with a silent Nvidia based graphics card because I just *HAD* it with ATI on Linux. My AMD chipset motherboard also had a lot of SATA instability under Linux and I had all kinds of problems letting the system know how to read any of the CPU's censors (X2 Phenom based CPU). So I have just ordered an Intel based CPU/chipset as well.

      I've no doubt that AMD is slowly

      • by Kjella (173770) on Thursday January 07, 2010 @05:50PM (#30688030) Homepage

        How many years was it again that they promised to produce open source graphic drivers for Linux?

        Announced: September 7th, 2007: press release [amd.com]

        Since then they've been catching up more and more, the HD5xxx/Evergreen/R800 instruction set was posted before Christmas so the docs are almost up to date, minus a few things like UVD2. Also AMD promised to help the open source community, not write the whole thing themselves and it's making big strides but there's also a lot of rework going on in xorg to support a modern desktop.

      • Are you that guy from the Debian mailing list? You sound like him - unjustly bashing ATI/AMD, misrepresenting their statements, and exaggerating the problems ATI on Linux has.

        They fact is that even the binary drivers (yuck) are much better than thy used to be, and the Free drivers are moving along by leaps and bounds. AMD has done very well with their promise to deliver documentation, and the Xorg guys are improving drivers as fast as they can, given limited manpower, and a rather large amount of (needed) c

        • by owlstead (636356)

          Are you that guy from the Debian mailing list? You sound like him - unjustly bashing ATI/AMD, misrepresenting their statements, and exaggerating the problems ATI on Linux has.

          No, for a normal user they are unusable. Any less advanced person would not have spend those kind of hours on configuring a graphics card.

          They fact is that even the binary drivers (yuck) are much better than thy used to be, and the Free drivers are moving along by leaps and bounds. AMD has done very well with their promise to deliver documentation, and the Xorg guys are improving drivers as fast as they can, given limited manpower, and a rather large amount of (needed) churn in Xorg (DRI2, KMS, TTM/GEM, Gallium3D) that they need to keep up with.

          Oh, I'm not bashing anyone, trust me on this.

          I'm just this guy waiting on some kind of normal display/sound drivers on my Linux computers. Currently doing anything slightly over running vesa or nvidia for graphics and very basic sound stuff sucks on Linux (or at least the last 4 Ubuntu versions I tried. Don't mistake this comment for "Linux sucks". I love the way many thi

          • >No, for a normal user they are unusable. Any less advanced person would not have spend those kind of hours on configuring a graphics card.

            I don't remember doing much configuration for my r300

            • by owlstead (636356)

              Well, congratulations for getting a well working configuration. But don't assume your easy configuration is the norm. Things are getting better, but we're not there yet.

              • All I did was pop in a Fedora 12 livecd and my R500 card started working. Absolutely no configuration. Whatsoever.

                Hell, my old roommate uses Gentoo and even he doesn't have to do much of any configuration to get it running, all he does is build X as usual, with radeon support. If you still need to do manual configuration of X on a modern setup, you are failing hard.

      • by lewiscr (3314)

        since my first slackware CD's

        Slackware comes on a CD now? I can finally get rid of my pallet of floppies!

    • Re: (Score:3, Insightful)

      by Sir_Lewk (967686)

      At first glance, from the subject line, I thought your post was a snide comment about the state of official ATI drivers on linux. I must say though, you guys are doing an excellent job at picking up ATI's slack.

    • Not fast enough, I dumped my perfectly fine Radeo 4850 in favor of a somewhat slower NVidia, the reason was that X support was hit and miss, half the 3d functions crashed X others worked. I then dropped in my NVidia card and everything worked out of the box.
      I do not care for how many years we got promises, the linux drivers suck donkey balls, and probably will be forever.
      Wake me up when the stability is up to NVidias offerings, or shock the Intel opensource drivers.

    • by Ash-Fox (726320)

      Support in the open-source drivers is being written as fast as ATI can verify and declassify docs.

      Personally, I've not been impressed with the 'correct' opensource effort when it comes to 3D acceleration support, see my blog entry [livejournal.com] for more details.

      • Re: (Score:3, Insightful)

        Your post is roughly fourteen months out-of-date. In the past year, TTM and GEM have both matured and been submitted to the main kernel, providing memory management services to nouveau, radeon, intel, and via. GLX 1.4 support is now advertised server-side for DRI2 stacks.

        In Mesa, most of GLSL is now supported by the drivers that can accelerate it, and the actual GLSL hooks are now in place for r600 and i965. Additionally, in Gallium, work is underway to provide GL 2.0+ on i915, i965, r300+, and nv30+ (all G

    • ATI at it again... (Score:2, Informative)

      by GooberToo (74388)

      Just upgraded my brother's laptop over the holiday. Seems ATI dropped support for his GPU in their proprietary driver so now he has a choice. Option one, use the open source drivers which provide no 3d acceleration. Basically 3D is completely unusable. Option two, use an older distribution which has the required version of X, kernel support, and all dependent software. And with the second option comes all the associated security issues of running an old and unsupported distro. He chose to run a current dist

      • >Seems ATI dropped support for his GPU in their proprietary driver so now he has a choice. Option one, use the open source drivers which provide no 3d acceleration.

        Bullpucky. Any/all cards that are not supported by the binary drivers do have 3D support from the OSS drivers.

        >For Linux there is still only one 3D option - NVIDIA. Period.

        Funny, my experience with 3D with both Intel and ATI has been great

        • Re: (Score:2, Interesting)

          by Raptor851 (1557585)
          Well...he does have a point for some cards..take r500 series for example (such as x1550). Proprietary drivers dropped support for drivers >9.3, radeon opensource drivers get an average of 10fps in older games such as UT2004 or less powerfull games like Touhou 8. radeonhd drivers work, but aren't much faster and are still fairly unstable (15-20fps average, crashing every 10 minutes or so, driver has a long way to go still as the game is still more or less unplayable). Note that this is on a card roughly
          • >[r500] radeon opensource drivers get an average of 10fps in older games such as UT2004 or less powerfull games like Touhou 8.

            Huh. I admit I have no personal experience with 3D OSS on the r500, I just knew support existed. Obviously not that helpful if Imperishable Night only gets 10fps, though.

            I hope r300g and r600g get usable soonish.

          • Oddly enough, I just saw that the RadeonProgram wiki page says TH08 is "platinum" with Mesa 7.5 on r500. Something doesn't seem right here...although I guess technically the definition of "platinum" they give doesn't say anything about speed, just correctness. Still, you might think that speed so bad that it is unplayable on low settings might be worth noting.

            • Not feeling like replying to the troll posts, but I'll reply to you.

              I probably reported the TH08 status; I'm kind of a danmaku fan. It was totally playable on an X1950 on an all-classic setup, but there's been a few regressions since and I bet the performance has dropped a little bit. We had to sacrifice a bit of performance to add stuff like FBOs and DRI2. It should start getting better soon though; we've started paying attention to speed and such so the next couple months should see big speedups.

        • by GooberToo (74388)

          All of the 3d games he had on his laptop are now completely unplayable; measured in fractions of frames per second.

          Bullpucky. Any/all cards that are not supported by the binary drivers do have 3D support from the OSS drivers.

          That's simply not even close to being true. Does my original quote sound like the open source drivers are providing 3D acceleration. Proprietary driver 20-70 fps depending on the game. Open source driver, unchanged settings, 0.008 fps. I rounded down the time in the provided fps for the open source driver, which make that number even larger than it actually is. In one of the games it took over two minutes to render one frame. No joke either. Does the later of the two numbers hint to you that the driver is n

HOST SYSTEM NOT RESPONDING, PROBABLY DOWN. DO YOU WANT TO WAIT? (Y/N)

Working...