Rendering Mike Pan's BMW1 benchmark under 10 seconds in Blender Cycles 2.68

Hi all,

After some playing around I was able to render Mike Pan’s BMW1 benchmark under 10 seconds. I was not recording the rendering process, because it slows down the system, while grabbing the screen.

with friendly regards


MB Asus Rampage III Extreme
Intel i7-920 @ 2.66MHz
6GB RAM DDR3 in triple channel (3x2GB)
1500W + 1200W PSU

Graphic cards:
NVidia GeForce GTX 580 + 3x 590 (overclocked)

Here is a proof:

and a little how to :wink:

all the new benchmarks will be posted here, in a day or two.

Short answer: use 7 GPUs. Nothing to see here, folks… :cool:

Hi, you probably missed, what influences the scene render time - the tiles order, dynamic/static BVH, and hiding the render update … and a little overclocking.

EDIT: using these 3 setting I have got about 12% improvement, and another 7% by overclocking.
EDIT 2: my 256x256 render time without other tweaks was 11s 52ms

There may be hidden pearls that are not so obvious on the first look :wink:

I would say that technically you’re not rendering the benchmark in under 10s, you’re rendering a tweaked version of it. :confused: The whole point of benchmarks is consistency, and when you twiddle the settings, all relative timing goes out the window. Many people would also report much smaller benchmark times if they were allowed to violate the “Change no settings” part of the benchmark.

Having said that, all those cards don’t seem to scale terribly well in any case, possibly due to the initial overhead and small scene, 'cause my single (overclocked) GTX 580 got these times:

Benchmark default settings: 63.2 seconds
256x256 tiles: 31.71 seconds
Your settings: 31.67 seconds

That’s with the browser running and Steam installing games in the background. At least for me, the other changes you made don’t make any difference whatsoever once I’ve already changed the tile size. That’s on a GTX 580 Lightning, running at 900MHz with 2500MHz memory.

So, my renders took 3.3x longer than yours, but you have 7x the GPUs installed. I think that the small scene may be affecting things and that for more complex / higher resolution renders you may notice better scaling, though, due to the initial overhead being amortized over a greater amount of time.

To be clear, this isn’t to say that I’m not jealous. I’d love to have a few more GPUs for rendering. :smiley:

Yeh, while it’s good for you all I see is epeen…
“Look at how amazing my computer is”

And I wouldn’t bother posting your results in the thread, like pointed out it’s not the idea of the benchmark to be the fastest, but to deliver the results with the given settings so people can compare.