Gamefest 2007 slides ?

[AlNom]

Have they just not been released? I'm particularly interested in *gasp* the graphics portion. Abstracts

• Picture Perfect: Gamma through the Rendering Pipeline
• Xbox 360 GPU Utilization: Past, Present, and Future
• GPU Font/Vector Rendering and Approximating Catmull-Clark Subdivision Surfaces with Bicubic Patches
• GPU Data Structures and Advanced Lighting: State-of-the-Art Techniques from Microsoft Research
• From the Trenches: Xbox 360 Development War Stories
• Performance Considerations for Graphics on Windows
• Windows to Reality: Getting the Most out of Direct3D 10 Graphics in Your Games
• Mapping the Dark Corners: Creating a Flexible Framework for Dynamic Shadowing
• Windows Vista Graphics Development Drilldown: Direct3D 10 and 10.1
• Xbox 360 GPU Performance Update
• Advanced Xbox 360 Graphics Techniques Using Command Buffers and Predicated Tiling
• Graphics Futures: Direct3D 11 and Beyond

*sigh*...

I realize they mention their release in mid-september... but... ack.. I'm impatient.

 

[Rys]

I asked Microsoft about these slides the day after they were presented at Gamefest, and they said they'd be posted soon, and that turned into mid-September as noted. Will ask again, since that entire deck set is worth a read.

I covered one of the presentations in my Develop 2007 AMD Tech Day coverage, FWIW (Bruce's talk).

 

[AlNom]

Awesome. Thanks.

 

[Zeross]

Rys did you have news on when the slides will be uploaded ?

 

[pipo]

http://letskilldave.com/archive/200...s-and-audio-recordings-well-most-of-them.aspx

A *ton* of stuff...

 

[kyetech]

what about these slides?

Graphics Futures: Direct3D 11 and Beyond

I couldnt see them.

 

[Demirug]

Some more and now on the official page: http://xnagamefest.com/presentations.htm

The “Graphics Futures: Direct3D 11 and Beyond” slides will not downloadable at all. But the talk (I was there) was not very specific at all. It was more general speaking about what may come in future Direct3D versions.

 

[kyetech]

Some more and now on the official page: http://xnagamefest.com/presentations.htm

The “Graphics Futures: Direct3D 11 and Beyond” slides will not downloadable at all. But the talk (I was there) was not very specific at all. It was more general speaking about what may come in future Direct3D versions.

Was there nothing juicy for the graphics whores to get excited about?

 

[Demirug]

At least not for me. Everything sounds like evolution and not like revolution. Additional without gaveling a tech preview of a Direct3D version you never know what’s in. They remove many things from the D3D 10 drafts over the time. It would be a surprise if this will not happen with D3D 11 again.

 

[kyetech]

ok lastly is there even a projected time frame for this... are we talking DX11 in around the early 2010 mark, that would put it at time frame where it could make it in to consoles for the 2011 - 12 area.

OT what sort of transistor budget and chip configuration could we expect then? and will intel change the landscape of that with their work circa 2009 onwards?

I spose that would be the million dollar question right?

 

[Demirug]

I haven’t seen a roadmap for DX 11 yet. But with the new “no caps only tech levels” rule I believe we see DX 11 together with the full next generation GPUs.

 

[nAo]

I believe we see DX 11 together with the full next generation GPUs.

That's right

 

[kyetech]

I haven’t seen a roadmap for DX 11 yet. But with the new “no caps only tech levels” rule I believe we see DX 11 together with the full next generation GPUs.

So early to mid 2010 could be about right then.


<completely off topic>

Do we think the likes of Unreal Engine 4 be based on DX11 or a mature DX10?
What do we think the next gen Consoles will base their tech on?

I would love to hear from developers about what type of rendering features will be kicking about in the DX11 / UE4 time era. (from the pov of naming techniques like area lighting / soft shadows, FS / object based motion blur, GI, dare I say raytracing for seconday effects (not likely I know)) But then also with disscussion on what the techniques would involve, EG: How motion blur in next gen would differ from this gen in visual appearance and computational complexity)

It would also be interesting to discuss where we think image quality will be in the PC space in 2011, will we be using something other than anistropic filtering by then? and how far does anti aliasing actully need to go before it could be considered solved?

Is that something you guys would be happy to start a thread on just for speculative fun? Or do you feel there really isnt enough to go yet to even bother?

</completely off topic>

 

[Demirug]

D3D10.1 will be part of SP1 which should be out early 2008. Therefore IMHO a D3D 11 is already possible in 2009.

As Epic is already working on Unreal Engine 4 I would expect D3D10. These guys are not known for riding the last technologies wave. But depending on the API difference custom versions of the engine that use D3D 11 seems possible.

IMHO new technologies will be more a question of shader power (and CPU power, too) that can be wasted (or invested ) than the API that is used. New APIs can help if they allow building the same effect in a cheaper way but without a proper base power a new API will not help much there.

 

[JHoxley]

From a few chats with the D3D guys over in Redmond I get the impression (note: not quoting, just my interpretation) that they're really focusing on usability ("RAD" if you like buzzwords).

Complex, realistic and high performance graphics are getting harder to achieve. Sure, we can re-use techniques and so on, but new technology is likely to focus on making it easier for content creators and application developers to put the pieces together quicker and easier.

Consequently my bet is on more universal effects - maybe a bit like Carmack did with the D3 engines and it's unified lighting architecture.

I dunno though, wait and see I guess...
Jack

 

[nAo]

We are going to see new features as well, brand new stuff, some expected..some not.

 

[Demirug]

We are going to see new features as well, brand new stuff, some expected..some not.

As always when we got a new Direct3D version.

 

[Andrew Lauritzen]

We are going to see new features as well, brand new stuff, some expected..some not.

MS seems pretty concerned about order-independent transparency lately, so I wouldn't be surprised if we saw some API/hardware features to that end (multiple Z-buffers for depth peeling maybe?). I'd be quite surprised if they "convince" AMD/NVIDIA to do something like a full A-buffer implementation in the near future though...

 

[Arun]

I'd be quite surprised if they "convince" AMD/NVIDIA to do something like a full A-buffer implementation in the near future though...

Pfff, I'm sure NV execs realize how much money they can save by going for a full TBDR next-gen design: they just need to say that to the ex-Gigapixel employees while mentioning they "aren't completely sure it makes financial sense yet", and they'll work 100 hours weeks for free during the next 3 years.

Wait, it doesn't work like that? Ah well! Point remains that I can't help but think that whenever these companies talk of A-Buffers, they really mean hiding that implementation detail under a TBDR-like architecture. Brute force only makes sense when the relative efficiency isn't so incredibly different...

 

[Demirug]

After making the MSAA buffers more accessible for the programmers I don’t think we will see another change in this area that soon.