Machine learning - Services for Research
CSC's computing infrastructure supports a wide variety of tools for machine learning and deep learning. In particular, compute nodes with dedicated GPU accelerator cards (Volta V100 and Ampere A100) are available.
Puhti and Mahti
A variety of deep and machine learning frameworks can be used in the Puhti and Mahti supercomputers, in particular the GPU partitions. Note that, as with any other tasks, the login nodes are not intended for heavy computing. Please use batch jobs instead. In particular, the login nodes do not have any GPU cards installed.
Several machine learning-related GPU-optimized Python packages can be found pre-installed, including scikit-learn, TensorFlow, Keras, PyTorch, and MXNet are available via the module system.
GPU flavors at cPouta and ePouta
The cloud services cPouta and ePouta contain several types (flavors) of virtual machines, including GPU accelerated ones with 1–4 GPUs attached. Virtual machines are particularly suited for cases with specific requirements for the operating system or software, or when administrative rights are needed. GPU-enabled Docker containers can also be run in GPU-flavor instances in cPouta and ePouta.