Submitted by “This morning at the X Developer Summit in the United Kingdom, Matthew Tippett and John Bridgman of AMD have announced that they will be releasing their ATI GPU specifications without any Non-Disclosure Agreements needed by the developers! In other words, their GPU specifications will be given to developers in the open. Therefore you shouldn’t need to worry about another R200 incident taking place. The 2D specifications will be released very soon and the 3D ones will follow shortly.”
This sound great. Hopefully some other hardware manufactures will follow suit.
Shouldn’t be any problems with the free software world now, as long as they don’t drag their feet with the 3D specs.
Well, there lies the problem. The R300 was meant to receive the ‘full specifications’ and every time the developers asked for more assistance and explanation, all they got in return from AMD was ‘you’ve got it all, piss off and stop whining’.
For me, I’m not going to believe it till I see a real working product that is either equal or better than the closed source one. If I am going to spend NZ$3K on a laptop, I want to be damn sure it’ll actually be fully supported rather than ‘kinda supported and really buggy’.
Also, I certainly don’t want to see a tonne of Linux’isms within the driver that’ll make porting it to *BSD’s and other UNIX’s more difficult than it needs to be. Open source application developers are already thinking the world revolves around Linux, lets hope that the driver developers have more common-sense than that.
Edited 2007-09-11 02:40
Kaiwai,
I’ve also been looking at laptops (available in New Zealand) that are great for software development for around NZ$3k (~US$2100). Check out the HP 8510w if you haven’t already. For the longest time this line of models (starting with the nw84x0 series) only had ATI graphics, which sucked on Linux, until now! Good on AMD/ATI! Due to the sucky support of ATI on Linux I was looking around at alternatives to this line of laptop for my next machine. But ATI’s announcement means they’re back in the running.
By the way, the 8510w is not now restricted to ATI only, you can now choose either NVIDIA Quadro FX 570M or ATI FireL V5600. You can’t lose.
Apologies everyone else for the first bit being only weakly related to the thread about the ATI spec release (disclaimer: I don’t work for HP or AMD/ATI, just wanna see that fellow coders are made aware of good kit). But I feel that this news from AMD might make a difference to people’s purchasing decisions (provided AMD actually deliver).
Now back to your usual programme …
Edited 2007-09-11 03:19
You both should check out http://lenovo.com .
The T-series is awesome. http://www.thinkwiki.org/wiki/ThinkWiki for more info.
Thanks for the link (although Lenovos are big bricks).
Fingers crossed Nvidia will pressured enough to release their specs – then we can write a decent driver for the Nvidia chip on the Playstation 3 (Ubuntu on PS3 runs a bit slow without decent graphics hardware acceleration).
Nvidia releasing specs to their “normal” chips won’t be enough for two reasons. The chip in the PS3 is a bit specific to the PS3, and second the hypervisor in the machine won’t let you access it.
I wouldn’t count on that till it is completely obsolete. Afterall getting royalities for access to the GPU is how Sony makes up for theis subsided PS3 prices. In some time they may release some limited sdk for the GPU for indie devs but that’s probably still far off.
Thanks for the link (although Lenovos are big bricks).
The T series are pretty slim and sturdy as hell. I got mine with Intel graphics and Intel wireless. It is fully supported under Linux.
I have a dv6209tx which is a Core 2 1.83Ghz with an nVidia go 7400 – basically it runs every operating system I’ve chucked at it. OpenSolaris runs nicely, along with Fedora, Ubuntu and SLED 10 SP1. With 2gigs of memory, you could say its a little pocket-rocket.
Its all basically going to hinge on what Intel does with their GPU designs and also what happens with the ‘gpu extensions’ ala SSE4 being promoted by AMD. Its difficult to envision where the future will lead but I hope that it’ll force all players in the market from relying on ‘secret sauce’ and instead compete based on technical merit and branding instead.
Linuxism is bad but *nixism is good ? lets go one step further and say *nixism is bad also and *ism is the good “ism” here yes ? no ? if no why ?
Code riddled with uniquely Linux’isms and GNU’isms are bad; name one thing out there which is riddled with UNIX’isms, or more correctly features which are ism’s that are not mentioned in the UNIX specification which Linux fails to implement.
I agree that those Linux’isms and GNU’isms are bad. It probably has to do with the fact that most applications are developed on GNU/Linux today so most developers don’t even know how to do POSIX or generic UNIX programming.
I am trying to make sure that all software I develop runs just as well on Linux, Solaris, BSD and Windows (if applicable) by avoiding all platform-dependent code as much as possible.
For example Gnash suffers (or suffered) from these a lot so it is (or was) really difficult to make it run on Solaris and BSD. With a little effort on the part of the main developers these issued could be resolved but as a Solaris or BSD developer you are welcome to point out the errors and submit patches for them.
I’m gradually learning C, thats one of the things I’ll force myself into getting my head around – that the world does not revolve around me, there are other people with other operating systems I should taken into consideration when writing an application.
At times there will be situations where I will have to reply on an operating system specific feature but hopefully I’ll try to create a platform independent library to work around that problem.
I’m quite happy to point out problems, the problem occurs when developers take these critiques as ‘personal attacks’ and fail to take patches on board which address these issues; wine and patches relating to Solaris support – for instance, is one example.
However, it isn’t a clear cut case – KDE are more than happy to welcome patches, be they for bugs or improve platform support – Solaris will be an official platform supported in 4.0 – which is great. I wish more projects were as enthusiastic about other platforms as KDE developers.
“Linuxism is bad but *nixism is good ?”
If your target platform is X then yes.
Because *nixism is portable, Linuxism is not.
Portable to what, *nixes ?
They have been making these sort of statements for years. Why is this any different? Why could they not release the specs at the X Summit? It would have been great PR and eliminated some of the scepticism.
The OSS train has already left the station and ATI is still buying the ticket. Even if they do release the specs, Intel and NVidia are still years ahead of them in development and distribution.
“Even if they do release the specs, Intel and NVidia are still years ahead of them in development and distribution.”
not for long
If they live up to their claims, this could be a very pointed turn of events for linux/3d. If they open everything up, it should pave the way to getting some good 3d accell into the kernel or at least a nice xorg driver. nVidia has always been the way to go, but AMD could seriously mix that up.
Even for OEMS like Dell who are shipping liunx, inclusion of drivers would mean they would probably default to recommending AMD cards/chipsets in linux based boxes.
If this pans out, I for one will be very anxious to swap out my nvidia board for a new AMD board, so I won’t be buying another card until I see what direction this is going in.
Even for OEMS like Dell who are shipping liunx, inclusion of drivers would mean they would probably default to recommending AMD cards/chipsets in linux based boxes.
I’m not so sure of that. Intel by far has the most Linux support of any hardware vendor. From graphics cards to wireless cards to sound cards. Once AMD has their own sound and wireless chipsets that are fully open source supported then maybe they’ll be recommended as much as Intel based computers.
AMD are getting a lot of positive press with this promise but they have lied before and it will take time before they can be trusted again (after releasing documentations and/or human readable code for their chips) until then this is just marketing.
Actually, I think the foss world has so much to gain and so little to lose, that this will be blindly welcomed by all the most hardcore doubters. Admitedly, this community has quite a lot of hardcore doubters and not without reason.
All that happens if they try to pull a fast one though, is that ATI suffers again on Linux and Nvidia resumes it’s position as the GPU of choice (when Intel isn’t enough for you anyway).
Basically, it’s all good.
*sigh* that should say
“…welcomed by all BUT the most hardcore…”
I’m an idiot.
** My mistake **
Edited 2007-09-11 06:24
Don’t look a gift horse in the mouth. AMD can give, and they can take away.
Open and free technical documentation?
Woa. Like a dream coming true!
😉
Will these specification make it possible to write a driver so I can put a 3D card in my headless server, have all OpenGL calls redirected and computed on this server and the results returned over the network?
I am using VNC to connect to my server and everything is running there but I can’t run 3D stuff this way. It would be great if I could run 3D applications and games on the server and transparently display them on my laptop in an efficient way!
The only thing is that this protocol probably wouldn’t be called VNC (Virtual Network Computing) anymore but ONC (OpenGL Networking Computing).
It’s perfectly possible to run opengl with an exported X display as long as the driver supports it. As far as I’m aware nvidia & intel drivers already support this config but performance won’t be great though
What I’d like to see is the server generating a realtime h.264 stream – anti-aliasing 3d stuff would essentially be done for free by the codec, with a decent quad-core chip or gfx card capable of realtime h264.
3D games are really low-latency realtime case: you need to deliver a frame between two vblanks, which is a problem over the network (network latency, en/decoding, buffering frames, etc.), especially if you spend time on en/decoding (also many codecs have frame difference stuff which may require more frames present to encode).
Edited 2007-09-11 08:48
Will these specification make it possible to write a driver so I can put a 3D card in my headless server, have all OpenGL calls redirected and computed on this server and the results returned over the network?
You know how some people say that because X is 30 year old technology, we should dump it and start again?
Well, what you’re asking for is 30 year old extensible technology (X), and the not particularly new GLX extension (GLX, not Xgl, which is something different).
Next time you hear someone say we should dump X, tell them to hollow out the bone in their head to make room for a brain. 🙂
(I think I misunderstood your ‘server’ to mean ‘server’ in the X sense. Rather than do the 3D rendering on the application server, do it on a thin client with cheap 3D hardware on board. You’re still running your actual apps on the application server, and getting beautifully rendered stuff on the display. Dave Richards from Largo is doing precisely this => check out his blog linked from Planet GNOME.)
Edited 2007-09-11 07:32
Maybe I wasn’t too clear in my question. The last possibility you mentioned as in the GLX extension is one option. You run your applications on the (headless) application server and display the output on your client having a massive 3D card using the locally DRI/OpenGL accelerated X server.
The other possibility is to have a 3D card in your (headless) applications server, do all OpenGL computations there and return the results somehow to your relatively weak client device with some older 3D card unsuitable to playing modern games with the local DRI/OpenGL accelerated X server.
For instance I have a Dell C600 laptop with an Ati Rage 128 Pro graphics card and it works fine for normal things. But when I want to play newer games I would have to buy a new laptop since I can’t upgrade the graphics on this machine.
My suggestion is if I would put a modern AGP 3D card in my server (a repurposed old desktop machine) and run the 3D games and applications there and only display the output on my laptop in an efficient way, would it work so I wouldn’t have to buy a newer laptop just because of the old graphics card.
Edit: I am probably thinking of what the Fusion project is going to with the integration of CPU and GPU, but that’s still a few years away.
Edited 2007-09-11 09:51 UTC
You’re optimising for the wrong thing. Sending relatively tiny GL commands and textures to a cheap, 3D-enabled thin client is going to give you a much better experience (and bang for buck) than trying to pump a very rapidly changing framebuffer across the network.
You mention “massive 3D card”, which is where I think your thought process goes wrong. It’s just not the case now that these things are inappropriate or too expensive for thin client hardware.
This is not a “plug and play” solution but if you have a bit of patience you could try using the chromium project:
http://chromium.sourceforge.net/doc/index.html
I used it for a while in combination with Xdmx to power my main workstation setup (6 monitors hooked up to 4 computers), but the configuration was non trivial. I tore that system down in favor of a single workstation with larger monitors and compiz. While it was up the system did do GL acceleration across the entire wall though.
I bet nvidia is feeling a bit of presure right now
Trying to figure out what goes wrong and where between your rendering code and the end result on the screen can be quite difficult at times.
For instance it’s not unheard of for a machine to BSOD randomly with some development versions of games, just because of a specific usage pattern of the 3d api that expose a latent bug in the driver – and you have no easy way of figuring it out since the drivers are essentially a black box.
I know that some kind of catastrophic event will have to happen for the gaming industry to start doing any significant amount of linux development, but things like this can help turn it into a more comfortable platform to develop games for than windows.
Edited 2007-09-10 22:41
Next months news:
> AMD open source driver does 3D, and does it well
> World of Warcraft ported to Linux
> Linux achieves world domination
Horay!
Edited 2007-09-11 00:27
Next months news:
> AMD open source driver does 3D, and does it well
> World of Warcraft ported to Linux
> Linux achieves world domination
WOW already runs fine on Wine in Linux.
WOW already runs fine on Wine in Linux.
WoW does not run anywhere close to *fine* on Wine. It runs. It gives ok FPS. But thats it. WoW is the kind of game you need to run very stable (Disconnecting while fighting bosses on Mount Hyjal is not popular).
70 Paladin, Linux gamer guild.
WOAH, it just keeps getting better and better. this means drivers are gonna happen and they are gonna be great. i cant wait.
this is really good news. once i see amd release the specs; I will upgrade to ATI. I hope Nvidia and other companies follow. It can be done, thanks to amd, they are showing it can be done.
You accidentally put AMD in there where you should have put ATI.
Used to be different companies, with different priorities and different strategies. ATI wasn’t aiming for making CPU+GPU chips. AMD is. AMD has far more reason for releasing specs than ATI ever had. I’m not saying we’re now guaranteed the perfect Free hardware treat, but don’t confuse ATI’s stupidity with AMD’s management. Two different beasts.
True, but at the same time AMD has done little to distance itself from former ATI decisions – no speech to announce a change in management.
But at the same time – I ask, where is the source, the documentation? where is the website for all this to occur? where is the announcement on Xorg for collaboration. Look at what Intel did and compare it to AMD.
I know I’m sounding jaded and bitter but I can’t stand companies who promise and fail to deliver, then fanboys coming out making all sorts of excuses as to why management at the company doesn’t show an ounce of leadership.
The community will lynch AMD/ATI if they don’t follow through on this. Trust me. I think it would be nasty to say the least.
AMD hasn’t done much to distance itself? Give them a break (and I rarely give ANY corporation a break!). They merged fairly recently. Two large companies, with some similar and many different markets.
AMD IS doing something different – we’ve asked for this FOR YEARS from ATI – a year after merging, they announce that they’re going to do more than NVidia – not make a great driver. Release what we need to make a completely open driver.
And still people complain.
Truly, you can NEVER please everyone.
Nice movement, IBM joins OpenOffice.org
http://www.openoffice.org/press/ibm_press_release.html
And iso certified of course 🙂
Edited 2007-09-11 06:21
i have always bought nVidia in the past because the supplied the best nix drivers……… that may change now.
So AMD/ATI are trying to fix their binary driver (and the results should be “coming soon”), as well as actually opening up specs. If they actually follow through, my next video card purchase goes to them (and I am in the market for an upgrade pretty soon)
This is really quite stellar news; it seemed like a pipe dream not so long ago.
The earlier announcement about cooperation with Novell was rather unappealing in that it sounded as if the binary driver was still going to play an important role, and access to information was going to be guarded by NDA. All in all, it didn’t sound like much to get excited about. However this new announcement has the potential to change the FOSS landscape dramatically.
All that remains is for AMD to follow through; and chances are they will do just that. Okay there will still be a lot of hard work for developers to create usable drivers. But at least they won’t be crippled by lack of proper information.
If this all comes together without a fuss, AMD/ATI will get a lot of praise and some business from me. But I won’t forget Intel for being there first.
I don’t see how this is news when it was announced several days ago. Also what most posts fail to mention is that AMD is not only releasing the specs but their developers that are currently working on the linux driver ( there are only a few people ) will continue to do so but will help the community as well. So the writing of the drive is not up to the Xorg devs alone and it should come out fairly soon. So if all goes as planned you might be able to run beryl out of the box with ATI hardware sooner than expected. I can’t even start to explain how happy this makes me since the only way I can change my video card is to replace the whole notebook.
Anyway looks like AMD is gunning for the *nix community crown and that’s alright with me. After all they already have an OS X driver that works really well …
I think the real sign in this is a hardware vendor is seeing the value of getting the hardware out the door with little to no barriers to it’s consumers! I think that ATI might have made a huge turn around in it’s Linux support initiative and I think this will make laptops, desktops, gaming platforms with ATI the single most desirable solution in the open source world. As our buying power grows, then our impact will be felt, more than it’s obviously has.
If AMD delivers the required specs and does its best to help the X.org developers deliver stable open drivers, I’ll (slowly) replace all the graphics cards on all my workstations (~10 of them) to AMD.
Don’t get me wrong, I’m not an FSF zealot – I truly appreciate nVidia’s work on getting a good stable driver for Linux/BSD/Solaris. But never the less, being closed source *, nVidia simply cannot use the community to solve long lasting deficiencies/problems. (Xen support under Linux/Solaris; 64bit FreeBSD support ,etc)
– Gilboa
* In nVidia’s defense, their XFree 3.x drivers were open; but C&D letters from Intel and Microsoft forced them to close the driver. (Sadly enough Novel removed one of the links… But I did find the second one)
http://marc.info/?l=dri-devel&m=114981283225530&w=2