R600 Gallium3D Shader Compiler Milestone Hit

[V!NCENT]

// Offtopic
Yo just a quick offtopic question (Googling did not do me any good, same goes for Wikipedia):

My 4870x2 died just before xmass (r600). I send it with RMA and had to use my onboard HD 3300 (r600). Now I got back a mail "Too bad. Pick something around 140 euro's) and got the 5770 (V3, dunno if that's marketing BS or if it actually means something...).

What core is the 5770? Is it r700? I got as far as the codename for the core, which was juniper, or jupiter... something like that
// Ontopic

[pingufunkybeat]

HD5xxx cards are Evergreen/r800.

Support in the open drivers is extremely basic at the moment

[droidhacker]

It might be best to leave it on the shelf a while and just keep using the r600.

[pingufunkybeat]

Well, he can use it with Catalyst, and check the free drivers from time to time and report bugs...

[droidhacker]

Well, he can use it with Catalyst, and check the free drivers from time to time and report bugs...

Can... but who'd actually WANT to?

[V!NCENT]

HD5xxx cards are Evergreen/r800.

Support in the open drivers is extremely basic at the moment

You have GOT to be kidding me! >.<

And just when open source was starting to get somewhere...

How can I use the 3300 when the 5770 is plugged in? I plan to use two installs of linux; one older distro with fglrx for the 5770 and a newer one with mesa for the 3300.

FileSystemChecK!!!

[V!NCENT]

Well, he can use it with Catalyst, and check the free drivers from time to time and report bugs...

Good idea! The sooner I can get my hands on a working FLOSS driver for my graphics card, the better...

I'm getting a bit tired of waiting for it, so how can I get hands dirty?

I don't have a lot of spare time left... Getting money is starting to take up too much time...

Card should arive between 2-5 working days...

I only know about a lot of things on the surface, but I'm willing to learn and go deep eventualy.

[droidhacker]

I think I understand what you are trying to say here, but so far EVERYTHING you've said on this topic is SPECULATIVE.

I certainly wouldn't accept any of this as FACT -- not without some kind of BACKING -- that means that you need to CITE YOUR SOURCES for your statements, and that typically means identifying press releases or other official statements including this information.

It certainly would be NICE if development followed the path that you are suggesting. I'm sure that we can ALL agree on this.

Just don't present your dreams as facts. Please.

And also note: You SHOULD care about your credibility -- without credibility, nobody will take you seriously. I'm afraid to say it, but you come off as a bit of a crackpot.

I will now disassemble your statement;

"Sources"

sometimes there is no need for an source to know that there is no public spec for the UVD2 part.
because there is no SPEC for the UVD2 part.

It is *impossible* to develop hardware of such complexity without specifications. This statement is therefore FALSE.

It is also a misinterpretation of the request for sources -- the sources requested are *official statements* -- i.e. Bridgman said on (date) that (something would be done).

ATI=Evil thats true! my only hope is that AMD brings a new style of being a company.

My R600 says yes.

to know that there is no need for a source!

Again... crackpot statement / guessing. In order for you to make statements about what is happening WITHIN AMD, you need to present EVIDENCE, and that means you need to reference official statements made by AMD people.

in 6 monds the driver dev for R900 starts thats means AMD do not have the resources and time to bring the R600-r800 opensource part to an good level.

This suggests a lack of understanding about resource assignment within an organization. It also suggests that you lack the ability to "connect the dots". AMD is clearly building up the open source driver in preparation for FUSION processors. It is one thing for your GPU to require specialized binary drivers, it is quite another thing for your CPU to require them. Does it not strike you as interesting that FUSION parts will start to become available at about the same time when we can expect a more-or-less complete open source R800 G3D driver?

because amd needs nerly all devs for the new r900 because they wana sell new hardware..

Yes, and this means getting the R800 open source drivers FINISHED in time for FUSION.

the opensource part of R900 will be much better because the opensource driver starts dev from the beginning not at the end of live time.

It certainly makes sense to begin preparations for ultimate goals as early as possible, but this statement lacks support. You need to cite your sources!

R900 also means there will not be a opensource driver for a hardware desined for a cloused source driver thats means amd will splitt the acceleration unit and the crypto unit because they wana have opensource drivers.

First off, your grammar is terrible. This statement is VERY hard to read. Also, as far as I know, the "crypto" stuff is relevant only to the UVD components. The rest of the specs released for the GPU had to be approved not for crypto stuff, but rather for INTELLECTUAL PROPERTY issues, i.e. it may be an AMD GPU, but that doesn't mean that they didn't use information they obtained under an NDA from other organizations! Giving out this information could open AMD up to major legal problems if not done carefully.

YES the SAME ATI Guys DEV the R900 BUT not the DEVs FAIL in the past only the Management?!!!

You don't even know if all the "devs" even worked for ATI. In fact, you know that DEFINITELY some/many of the devs did NOT work for ATI, hence the IP issues related to release of specifications.

Yes, I agree that there was a major management screwup. They were very shortsighted in encumbering their hardware with all those NDAs. We can only HOPE that they have thought it through better for next time around.

AMD and R900 means the ATI devs do have a new Managment and they can make a Better product because the Managment chance the way of thinking abaut a good product.

It means that they have the CHANCE of doing better. It does NOT FORCE them to. They are still free to screw up badly.

ATI=Evil means the Managment wana have DRM(Copyprotection) then the DEVs built it (at all cost --> hurt opensource)...

DRM is another issue. It hurts a lot more than open source -- it hurts EVERYONE. But fact of the matter is that sometimes you have to just accept it. Problem here is that stupid windoze users want to be able to play the encrypted disks that they should never have accepted in the first place.
** Just note: If the regular consumer wasn't a complete retard, then the regular consumer wouldn't accept media content that contained DRM, which means that the media PROVIDERS wouldn't be able to sell ANY of it. If the media providers found that DRM made it impossible to sell media, they would sell media FREE OF DRM!!! The regular consumer windoze retard is to blame for this. That and the US-DMCA that makes it illegal to decrypt encrypted data. Now I ask you this: isn't the purpose of encryption to protect data you don't want decrypted? If that is the case, then why should it be protected by LAW? Legal protection should ONLY be available to NON-ENCRYPTED data. The act of encrypting your data should be considered the complete assumption of responsibility for protecting your data. By NOT encrypting your data, you choose NOT to protect your data, and again, this should not be protected by LAW.

In other words, the LAW has no business interfering with data.

AMD=Good means the managment wana have a opensource frendly hardware AND DRM(copyprotection) then the devs built it.

AMD has shown to be good, but our exposure to AMD has been through good people, like Bridgman. We really don't know what the top level execs are planning -- one thing we can be certain of though, at the top level, they are looking at one thing: the bottom line. You can be certain that they don't even know what open source IS. So what they choose to do remains to be seen.

there is no source needed to understand this little tiny different!

Of course there is! You have made statements based on WISHFUL THINKING and present them as FACT. They are NOT FACT and even fly in the face of everything logical. If you want to make a statement, you NEED TO BACK IT UP!!!

[bridgman]

For the record, I come from the ATI side and joined the company in 1998. If that makes me evil, so be it

[monraaf]

i do not care abaut my own credibility !
in 6 monds the driver dev for R900 starts thats means AMD do not have the resources and time to bring the R600-r800 opensource part to an good level.

because amd needs nerly all devs for the new r900 because they wana sell new hardware..

the opensource part of R900 will be much better because the opensource driver starts dev from the beginning not at the end of live time.

You live in a fantasy world. ATI is still fiddling around with code for r300 hardware, and by doing so has created a gigantic backlog. Evergreen hardware has been out for over 6 months and still hasn't got any acceleration code. Don't expect it to be any better for R900. Remember that AMD has only 2 devs allocated to the opensource drivers. This is going to take ages.