ATI's Fuddy, leaks and spin

[saweetnesstrev]

ATI's Fuddy, leaks and spin

UNDER THE HAND of Dave Baumann, Beyond3D increasingly establishes itself as one of the wibbly Webs most insightful and reliable resources on virtually all things 3D graphics related.

OK so there are a minority of ignorant, opinionated fools on its forums, who lack the elementary wit to understand the challenges of investigative journalism, and the delta between what a hardware enthusiast site does, and all our best efforts here at The INQ. This is despite their enjoying the fodder we offer them, but if you want the detailed low down on 3D graphics technology then Beyond3D is probably where it's at.

It is also one of ATI Technologies favourite hangouts and a place where we're told that even Nvidia employees ? contrary to their internal policies - post under pseudonyms.

This, we suggest, can present challenges to a site such as Beyond3D, which needs a close and cuddly relationship with manufacturers to secure early product intelligence and samples, but if there's evidence of inappropriate tampering on its forums to maintain those relationships, we sure haven't seen it yet.

Perhaps evidence of this, is the existence of a thread that has been running over the past couple of days, on a subject that ATI would love to have buried, and buried deep.

The hoo ha kicked off when a Microsoft Powerpoint presentation ? marked up as 'ATI Confidential' - which was composed by ex-Nvidia technical evangelist and now Codemafia Godfather, Richard Huddy, was, now claim ATi, inadvertently posted up on its website.

A link to this, and various 'Mirrors' ? some of which seem to have been pulled (at the request of ATI?) - and fevered commentary was posted in the Beyond3D forums.

For good reason, Richard Huddy is regularly rolled out by ATI as its European Developer Relations Manager and this was his Powerpoint presentation to developers at the recent Gamers Developers Conference in San Hose, California.

Unfortunately for ATI, and perhaps Richard, the presentation also included his confidential notes, and this has embarrassingly exposed not only an insight into how ATI tries to manoeuvre and pitch its audience, but also a number of interesting gems relating to its own, and Nvidia's forthcoming technology.

Perhaps one of the most interesting, was confirmation of one of our Fudo's, here.

. We're told that on page 31 of the document, Richard advises himself: ?Steer people away from flow control in ps3.0 because we expect it to hurt badly. [Also it?s the main extra feature on NV40 vs R420 so let?s discourage people from using it until R5xx shows up with decent performance...]

Fudo reported that the forthcoming ATI R420 won't feature Pixel Shader 3.0 and this is what the document seems to confirms, and further, that it is a weak point ? from a marketing perspective ? for ATI, as Nvidia's forthcoming NV40 definitely will have Pixel Shader 3.0 support.

What near-future practical benefit this will have to end users though is debatable. Because even when ATI and Nvidia introduce their respective next generation products, as we understand it Pixel Shader 3.0 games will be at least 6 ? 12 months down the road, and another generation of graphics marchitecture is likely to be upon us.

During our research for this article with games publishers, we were unable to conclusively determine whether any games that will feature PS 3.0 are under development. In fact some of the games publishers referred us back to the GPU manufacturers. Id Software's Doom III seems the most probable candidate but Activision was unsure whether an official announcement on this had been made.

We also asked numerous sources including both ATI and Nvidia whether a PS 1.1 or 2.0 game could be patched to support PS 3.0. However no-one we spoke to knew whether this was a task that could be undertaken, and only ATI suggested that it would probably be impractical.

That said, as sure as the chicken follows the egg, to deliver future PS 3.0 games, game development companies need the hardware to develop and test on. So in our view it would have made sense for ATI to have introduced it if it could have.

The fact that Richard mentions, the next-next generation product ?R5xx? and that this will have PS 3.0 capability, shows that ATI knows it's a desirable and relevant feature, but that it wants to ?steer away? and ?discourage people? because it ? ATi ? will ?hurt badly? when compared to Nvidia NV40.

Firstly, we wonder which people? ATI wants to "steer away" and "discourage"? As the target audience for the presentation was game developers, then it's reasonable to assume that these people are the gaming development community - and if so, this is atrocious.

The way this could come across is that that for its exclusive commercial gain, ATI wants to stifle game developers - and thus limit the end-user experience ? until, that is, ATI is in a position to capitalise finacially.

But there seems to another viewpoint on what ATi was saying here, and you'd need to love 3D graphics tech, or be paid to do so, to fully understand it.

Basically ? as i understand it, and frankly I fully don't ? but it's that in the context of the preceding presentation slides, and the particular slide that the notes appear on, when Huddy says ?Steer people away from flow control in ps3.0 because we expect it to hurt badly? it's suggested he is saying that if developers endeavour to do code branching in a GPU's Shader Units, then performance will be poor. Period.

However the killer part of what's said is that because the next, next generation ATI GPU, which he confirms is ?R5xx?, will have a totally different architecture, than both the near to release R420 and Nvidia's NV40, then that architecture will address the poor performance of doing branching in Shaders.

It seems that what is planned for ATI's 'R5xx'is a fully unified Shader architecture ? a single engine so to speak - as opposed to the current architectures which have independent Pixel and Vertex Shaders.

Apparently in one of its own GDC briefings, Nvidia itself has suggested that code branching in Shaders is presently best avoided.

Which viewpoint is correct, we're uncertain but either way we suspect that ATI wishes it had been rather more diligent in its activities.

We asked ATi for a comment on the document and were eventually forwarded to its Desktop Products Public Relations Manager, Ms. Patti Mikula.

We asked Patti not only whether she had any comment on the notes in the presentation but also whether, as at the time of our INQuiring, as this is so close to the publication of ATI's eagerly awaited fourth quarter financial report, if she would confirm whether the posting of this document was not an example of Selected Disclosure.

Here's what prompt Patti had to say

: ?Thanks for the e-mail. Many of the comments in the speaker's notes are easily misinterpreted outside of the context of the spoken presentation. However, because some of the comments relate to unannounced products we will not be providing further information or clarification at this time. None of the information in the notes or in the presentation is material, so the proximity to our financial reporting isn't relevant.?

The inadvertent release of this document to people who have not signed a gagging NDA is harmful to ATI. Whilst you might need to be David Kirk to fully understand it, and all the implications of what is said, we think the detail of the document gives specific insight into ATI's long term product strategy. ?

Source: http://www.theinquirer.net/?article=15106

[Nathanael]

Now that was an intresting read....

[Glowstick]

Regarding patching games for PS3.0... well, you should take a look at the FarCry 1.1 patch notes...

[Chode]

Wow. That really was a very interesting read