iray for exterior scenes

[pixelperfectg]

FWIW I'm working on a multi-post blog entry on using iray in 3ds Max on an architectural exterior scene. Part one is online now: http://jeffpatton.net/2011/06/iray-for-exterior-scenes/ and I'll try to get part two (textures) completed and up soon.

 

Hope this doesn't come across as advertising, that's not my intent. I just thought it might be useful info to share at an architectural-vis site. If it does come across as advertising/SPAM please let me know and I'll delete this post.

[matthewgriswold]

Hope this doesn't come across as advertising, that's not my intent. I just thought it might be useful info to share at an architectural-vis site. If it does come across as advertising/SPAM please let me know and I'll delete this post.

 

Ha, Jeff, no need for the disclaimer, at least in my eyes. Good blog post. It seems like you've really taken a liking to iRay, so keep it up and we'll keep reading.

[notamondayfan]

Its really interesting to see Iray being used in Arch Viz in a production type scenario.

 

What's your initial reaction to using Iray and GPU rendering? How does it compare to other renderers?

 

Also I guess the key question is how fast is GPU compared to CPU? I know they are hard to compare directly, but what are your thoughts on speed?

 

Looking forward to part two! Cheers, Dean

[pixelperfectg]

What's your initial reaction to using Iray and GPU rendering? How does it compare to other renderers?

Like most render engines it's not the right tool for every job out there. However, if you can squeeze your scenes onto a newer FERMI based GPU, it can be very fast & efficient.

 

Also I guess the key question is how fast is GPU compared to CPU? I know they are hard to compare directly, but what are your thoughts on speed?

I only have two workstations here and they each have single CPU's (Intel i7 960's @ 3.2GHz) and there is NO comparison for me. Just one of my GPUs will typically run circles around the CPU. If I use three or four GPU's then it's stupid fast even compared to the two workstations using DBR. Of course I would probably be less shocked by the GPU performance if I were using newer CPU's and/or multi-CPU systems.

 

I think I need to connect with some users running more elaborate/newer multi-CPU systems and setup a benchmark scene to compare results so I and others will have a better grasp on real-world CPU performance vs. GPU performance.