Questions for 3D Labs/Creative?

[vidar]

Hi!

Someone at Creative has agreed to answer some questions about the 3D Labs acquisition. Do you guys have any questions you want me to include?


Vidar
www.templeoftech.com

 

[Gubbi]

Yes.

Considering Creative bought 3Dlabs on the merits of their upcoming products (programmable everything GPU) and the fact that 3Dlabs traditionally has sold into the high-end graphics adapter market.

1.) Will 3Dlabs based Creative products be affordable by mortals (home users)
2.) Will Creative take over the driver team aswell (I hope, because traditionally Creative drivers are CRAP).
3.) What exactly are 3Dlabs creating (pixel/texel fill rate, Vertices/s, number of registers/constants in programmable pipeline etc)

Cheers
Gubbi

 

[mat]

Will they continue to sell NV Cards or will they switch to 3dLabs (if they start/resume producing Consumer chips)

Why didnt they just use NV's Quadro Chips?
(i think NV is not very happy about this... )

 

[Mephisto]

1) Does Creative Labs intend to scale down/over future 3d labs technology to the pc performance and pc mainstream segment or will it exclusivly focus on the professional Opengl market?

2) Your comments on the insolveny of the 3D Labs competitor ELSA AG?

3) Up to now, Creative labs doesn't offer 3d cards based on ATI or ST chipsets. Might Creative Labs offer cards with non-Nvidia chipsets in the future (beside 3dlabs tech of course)?

 

[MfA]

How does the quoted 170 GFLOP's for the new architecture break down? Is that all single precision and usable in programmable part of the architecture? Or is it mixed precision and/or divided between fixed and programmable parts of the architecture?

To what extent will the flexibility of the architecture be exposed? OpenGL 2.0 is a step forward, but it cannot expose "general purpose programmability" which the architecture is said to posess.

 

[pascal]

Some layman´s questions:
- Will it have 64bits precision?
- How many levels of multitexturing in an single pass?
- Any memory bandwith saving technology?

 

[Dave Baumann]

If people from other sites wish to fish for questions here then that fine, but we reserve the right to use them ourselves seeing as this has already been asked here

 

[Ty]

Why not STM?

 

[psurge]

To add to MfA's question: how much of this GFlop number is coming from vertex processing, and how much from pixel processing? Are these two types of processing even separate? Will the precision of rendering be comparable to their professional CAD cards? Will they have full
support for OpenGL2.0/2.1 (*)? What type of FSAA solution will they provide (specifically, will SuperScene AA or a variant thereof be used?)

Serge

(*) OT: does anyone know anything about OpenGL2.1, specifically
major features it has versus 2.0?

 

[Humus]

Is there such a thing as OpenGL 2.1 yet? Now even OpenGL 2.0 is exactly close to final state yet.

 

[phynicle]

long lost pizel fusion 150...is that really you????

 

[psurge]

Is there such a thing as OpenGL 2.1 yet? Now even OpenGL 2.0 is exactly close to final state yet.

It is mentioned once in the OpenGL2.0 docs (the memory managment paper I believe)...

 

[Humus]

Oki, found the reference to it, but it's just a "we can push this functionality unto 2.1" in case there isn't enough interest in having it already in 2.0. I think it's a little too early to speculate on OpenGL 2.1 features now