How to use Blender on Shamu to render your scenes or animations.

First you will need to upload your *.blend file to your home directory.

If you need more than 128GB of memory request access to the bigmem queue. Email your request to If less than 128GB of memory you can use the "all.q" queue. In the submit script below, if less than 128GB of memory is required change "#$ -q bigmem.q" to "#$ -q all.q".

If you do not already have one, request a /work/abc123 directory. Once created create a subdirectory called "blender" and place your .blend file and another other files needed for rendering.
[abc123@login ~] cd /work/abc123
[abc123@login ~] mkdir blender
[abc123@login ~] mv ~/*.blend /work/abc123/blender

Create a submit file and edit accordingly. The contents below are using the Blender benchmark demo.
#$ -cwd
#$ -q bigmem.q
#$ -pe threaded 100
. /etc/profile.d/
module load blender
cd /work/abc123/blender
blender -noaudio -b benchmark.blend -o ./output/ -s 1 -e 100 -t 100 -a

The blender command above "-s 1 -e 100" means start at frame 1 and end at frame 100. "-t 100" means to use 100 cores. "-a" means run until completion. You MUST use the same number with "-pe threaded XXX" as the "-t XXX value". If you do not you will potentially overload the node.

Now submit your job to the cluster:
qsub benchmark.qsub

Now you have a directory with a bunch of images but you wanted a video file. Not a problem! We can use command line tools to encode the images into a MP4 file playable on any modern computer.

From the login node we use a software tool called ffmpeg to convert the images. The instructions below take a set of images (named 0001.png, 0002.png, 0003.png, etc..) and encodes them using the H264 video codec. Edit accordingly for your images:
[abc123@login ~] cd /work/abc123/blender/output
[abc123@login ~] ffmpeg -i %04d.png -vcodec libx264 video.mp4

-- AdminUser - 10 Apr 2018
Topic revision: r2 - 11 Apr 2018, AdminUser
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding UTSA Research Support Group? Send feedback