Sheep It render farm try-out

Sheep It render farm is a community-driven render farm, where individuals render each others’ Blender frames.

I’ve been an avid contributor to this service, I have gathered more than 10,000,000 points by rendering other people’s frames.

Having so many points puts me first in the render queue if I render something, so I decided to make a small test.

One frame of the rendered image sequence

This is what I ended up with.

To render the same with only my local computer, it would take 
10 hours and 38 minutes on CPU or 
2 hours and 36 minutes on GPU.

On Sheep It it took 30 minutes.

350 of the 360 frames were rendered in 20 minutes
Some of those frames took only 36 seconds to render. It helps a lot to check also GPU rendering as “Compute method” in addition to CPU.

Most of the frames were quickly rendered. I had the GPU Compute method checked in addition to CPU

The next 6 frames took 5 minutes on top of that

Then I decided to Reset or Re-submit the last remaining 4 frames, that seemed to take too long to render. You can do it in the Status > Processing frames page.

Re-submitting or Resetting the frames that take too long to render

And so, after 30 minutes all the frames were done.

Downloading the finished frames starts up pretty much immediately, but the speed seems to be 500KB/s maximum, so it takes a while to get the frames.

All in all, Sheep It is a very handy service for small, not-too-secret projects, since the rendered frames can be seen publicly.

Here’s another test I made, I tried the splitting option to render each frame with 6x6 tiles, each tile with different rendering node. It worked beautifully.

It got fully rendered in 2 hours and 23 minutes. With local rendering by only one machine it would’ve taken 5 days, 9 hours and 10 minutes.

One important note, though. Your projects will be deleted automatically within 9 days. So, it’s important to grab your frames into a safe place and maybe copy the list of contributed renderers in case you want to attribute them later.

The project will be automatically deleted in 9 days.