Energy consumption evaluation of Blender's image renderer in HPC environment

In our contribution we evaluate the energy consumption optimization of image rendering on a typical architecture of an HPC system. We use the renderer CyclesPhi, which is our own modified version of the Cycles renderer from the Blender 3D creation suite. CyclesPhi fits the HPC environment in such a way that it runs as a client on one or multiple nodes, and efficiently utilizes the cluster through optimal load balancing. In order to reduce the energy consumption of a scene rendering we used MERIC, our own developed library for HPC application profiling and runtime tuning. MERIC searches for the configuration of hardware, system software, and application parameters which can provide minimal energy consumption for each manually instrumented region inside the analysed application. Thusly we instrumented the Blender client and analysed the rendering task. On Haswell architecture (two Intel Xeon E5-2680v3 processors per node) we were able to reduce energy consumption by 9% while extending the rendering time by 21%. If a less energy conservative setting was applied, we would save 4.8% of energy whilest only prolonging the rendering time by 4%.