Services

Massive GPU Cluster for Earth Observation (MAGEO)


To facilitate application of Deep Learning to Earth Observation data, NEODAAS have a large Graphical Processing Unit (GPU) cluster comprising of five NVIDIA DGX-1 MaxQ nodes connected to 0.5 PB of dedicated storage. NEODAAS are able to offer compute time development assistance with applying the latest Deep Learning techniques to EO data.

The Massive Graphical Processing Unit Cluster for Earth Observation (MAGEO) was funded through a NERC transformational capital bid in 2019 to provide NEODAAS with the capability to apply Deep Learning, and other algorithms benefiting from a large number of GPU cores, to EO data. The cluster is operated as a service with researchers able to make use of the compute power offered by MAGEO and the expertise of NEODAAS staff.

Please contact us if you are interested in using MAGEO.

MAGEO installed in server room

Above: MAGEO installed in the new data centre at Plymouth Marine Laboratory

Hardware

The cluster is built around 5 NVIDIA DGX-1 MaxQ nodes, providing a total of 204,800 CUDA cores. The special edition NVIDIA DGX-1 MaxQ nodes are a more energy efficient version using 50% of the power of the standard DGX-1 while still delivering 80 % of the capacity. Each DGX-1 has 40,960 CUDA cores, 40 Intel Xeon 2.2 GHz cores, 512 GB system memory.

The cluster has a dedicated 500 TB of fast storage as well as being able to access satellite data stored on the existing file system at PML which has over 3 PB of storage.

Software

MAGEO is built around the following core technologies:

  • Singularity - to provide environments with the latest Deep Learning and GPU optimised libraries (e.g. TensorFlow, PyTorch, RAPIDS).
  • SLURM - for controlling jobs
  • JupyterHub - providing an easy method of accessing MAGEO from within a browser

Training

An NVIDIA certified training program on Deep Learning for computer vision was held on the 12th March 2020 at Plymouth Marine Laboratory. A MAGEO ‘hackathon’ was planned for 2020 but is currently on hold due to COVID-19, we are aiming to hold a virtual event in the future. Please contact us if you are interested in these or other training opportunities.