Join the 155,000+ IMP followers

www.ptreview.co.uk

High-Performance Computing Architecture for Seismic Engineering and Energy Research

TotalEnergies, Dell Technologies, and NVIDIA have partnered to develop Pangea 5, a supercomputer designed to enhance subsurface imaging and optimize integrated power models.

  totalenergies.com
High-Performance Computing Architecture for Seismic Engineering and Energy Research

TotalEnergies has contracted Dell Technologies and NVIDIA to design and install Pangea 5, a high-performance computing (HPC) system located at the Jean Féger Scientific and Technical Center (CSTJF) in Pau, France. The system is engineered to provide a sixfold increase in the company’s existing computational capacity to support seismic engineering and research and development in artificial intelligence.

Technical Architecture and Responsibilities
The deployment utilizes a specialized hardware stack to manage complex industrial simulations. NVIDIA provides the core processing and networking architecture, utilizing a combination of Graphics Processing Units (GPUs) and Central Processing Units (CPUs) optimized for massively parallel computations. Communication between nodes is handled via the InfiniBand networking platform, which ensures the high bandwidth and low latency required for scientific workloads.

As the primary systems integrator, Dell Technologies oversees the design and installation of the infrastructure. This collaboration addresses the requirement for industrial automation in data processing, enabling the deployment of advanced algorithms for subsurface imaging. These tools are critical for enhancing the accuracy of geological models, which facilitates low-emission hydrocarbon exploration.

Energy Efficiency and Operational Integration
A primary technical requirement for Pangea 5 is the reduction of the environmental footprint associated with large-scale digital infrastructure. The system’s architecture achieves a 40% reduction in energy consumption compared to previous iterations at equivalent performance levels. Technical optimizations in the cooling subsystem result in a fivefold reduction in energy demand for thermal management.

The installation features a heat recovery mechanism integrated into the CSTJF facility’s infrastructure. Residual thermal energy generated by the processors is captured and redistributed to provide heating for the center’s buildings, which accommodate approximately 2,500 personnel.

Industrial Applications and Impact
The computational power of Pangea 5 is allocated to several high-demand technical areas:
  • Seismic Engineering: Accelerating the processing of high-resolution subsurface data to improve resource localization.
  • Integrated Power Models: Simulating complex energy systems to optimize distribution and transition strategies.
  • AI Development: Reducing training and inference times for machine learning models used in predictive maintenance and operational logistics.
Scheduled for commissioning in 2027, the €100 million investment represents a shift toward accelerated computing in the energy sector. By leveraging parallel processing, the partners aim to maintain process stability while managing the increasing data density inherent in modern energy transition projects.

Edited by Evgeny Churilov, Induportals Media - Adapted by AI.

www.totalenergies.com

  Ask For More Information…

LinkedIn
Pinterest

Join the 155,000+ IMP followers