Eni in Italy Unveils Petaflop HPC

Eni S.p.A., an oil and gas exploration and production company, has unveiled a Petaflop class supercomputer at its Green Data Center in Ferrera Erbognone, Pavia, Italy.

The supercomputer will support hydrocarbon exploration activities, providing the computing capacity needed to improve the accuracy and resolution of the geophysical and geological studies, which are essential to locate and develop new hydrocarbon reserves and to project safe and optimal drilling activities. Eni will have an available computing capacity in excess of 3 Petaflops. The system is able to achieve a performance comparable to most powerful computing centres WW.

The supercomputer is based on a cluster architecture with a system consisting of 1,500 dual CPU nodes (24,000 cores), enhanced by 1,300 General Purpose Graphics Processing Units (GPGPU), which are used as computational accelerators (equivalent to 20.800 additional cores). Cluster nodes are connected by a high speed internal network and storage is assisted by a high bandwidth parallel storage system providing 5PB disk capacity.

The computing capacity will be maintained to the level needed to match business requirements with an annual replacement of half of the nodes every year in order to keep the pace of the HPC technology development cycle. The system will support the execution of the Eni leading edge suite of 3D seismic imaging packages, as well as advanced petroleum system modelling Algorithms.

In line with the Eni commitment to sustainability, the system is designed to target the maximum level of energy efficiency. For this reason, the supercomputer has been hosted in the new Eni data centre, the ‘Green Data Centre’ in Ferrera Erbognone (Pavia). This new centre has been built to host Eni’s central computer processing systems, both for information management and seismic simulation processing (HPC).

The Green Data Centre uses innovative infrastructure for energy efficiency, cutting CO² emissions by 335,000 tons per year (about 1% of the Italian Kyoto energy target) and reduces operating costs.