Mandelbulb in unlikely places

[cbuchner1]

Here a rendering of a Mandelbulb (probably the most non-commercial object ever) is used to promote a technology conference and in particular an emerging companies summit related to parallel computing with CUDA.

http://www.nvidia.com/object/emerging_companies_summit.html

Uh, was this particular rendering taken from David Makin? Is its use okay in this context?

[David Makin]

Uh, was this particular rendering taken from David Makin? Is its use okay in this context?

Not one of mine

[Buddhi]

It looks like Daniel White's image. It is his style of rendering.

[lycium]

It looks like Daniel White's image. It is his style of rendering.

Daniel's images are excellent, but it's a bit of a stretch to say that direct illumination from a coloured environment is his style of rendering 

GPUs are very well suited to rendering mandelbulbs, so it's actually highly likely that such an application would be mentioned on nV's page, especially right after the (paper) release of their supercomputing chip, Fermi

[twinbee]

Er yep, that one's mine I think

They didn't get permission, but one can't automatically blame Nvidia since it's possible they just bought it from someone who ripped the picture and put it on an iStockPhoto type site.

It's pretty funny, because I'm only using the raw CPU, no GFX card anyway atm.

especially right after the (paper) release of their supercomputing chip, Fermi

What do you think will be more suited to fast flexible raytracing (either GI or DI), the upcoming Fermi/GF100, or Intel's Larrabee?

[Buddhi]

In my opinion if they want to use you image for their commercials, they have to pay you or at least have your permission. They makes money with using your image. No matter from where they got this image.

[hobold]

Larrabee is no longer quite so upcoming, unfortunately. I think Larrabee would be a better architecture for ray tracers when you aim for more sophisticated simulation of light, or more sophisticated/complex data structures and algorithms. Fermi, on the other hand, would have more raw power, and would probably outrun Larrabee whenever you can apply brute force or some approximating cheats.

Both are interesting approaches to massively data parallel programming. But in my not so humble opinion, both these chips are trying to fly before they can run. We still haven't gotten all the fundamentals of the SIMD (Single Instruction Multiple Data) paradigm right.

Edit: had omitted one word, thus losing the meaning.

[KRAFTWERK]

In my opinion if they want to use you image for their commercials, they have to pay you or at least have your permission. They makes money with using your image. No matter from where they got this image.

I agree, You should definitely get something for this Twinbee!, a GFX card maybe?
They have no right to use it without your permission!

J

[cbuchner1]

That picture is worth a GTX 480, no less.

[lycium]

What do you think will be more suited to fast flexible raytracing (either GI or DI), the upcoming Fermi/GF100, or Intel's Larrabee?

possibly larrabee, but that's vapourware still. fermi has been paper-released, and apparently you'll be able to buy one in early april it's a monster, not necessarily for gaming (radeon 5000 series has that covered), but for general computing it looks to be awesome!

have a look: http://anandtech.com/video/showdoc.aspx?i=3783&p=6