We are happy to announce that the first GPU node went into production in Göttingen today. The node (ggpu01) has the following hardware:
- 2 x Intel Skylake Gold 6148 CPUs (40 cores per node)
- 192 GB memory
- 1 x 480 GB solid-state disk
- 1 x Intel Omni-Path host fabric adapter
- 4 x NVIDIA Tesla V100 GPUs
This node will be a shared node for data pre- and postprocessing as well as machine learning. The CUDA and cuDNN libraries are installed on the node and a GPU accelerated TensorFlow environment has been added to the anaconda3/2019.03 module. More information on how to use the GPUs can be found in our documentation.