Advances in the processing and analysis of seismic data, together with reservoir modeling and simulation, have been closely tied to developments in computing technology. Parallel processing and storage have enabled successive generations of seismic imaging and reservoir simulation techniques, together with increased scalability and reduced cost.
Currently most HPC systems used in the oil and gas industry are clusters based on x86 architectures, most components being standard technology from established manufacturers. Some of these installations are very large, with over 50,000 nodes installed in some major seismic processing centers.
The challenge now is to increase overall HPC throughput and productivity while overcoming some physical limits, whether these are electrical power, heat dissipation, footprint, cost or a combination of all of these factors. Likewise performance and scalability depend on the optimization of the application software to exploit hardware scalability and to generate faster turnaround times.
For these reasons the oil and gas industry is looking closely at the potential use of accelerator technologies to speed up certain parts of the application code which are suitable for parallelization. ‘Many-core’ technologies show strong potential for accelerating these parts of the code, but require programming changes to enable this. The skills and knowledge needed to achieve this can be obtained from a suitably-experienced HPC system vendor, either through a consulting and training agreement or as a turnkey solution.
Customer Needs Summary
Improving seismic data and reservoir model quality
Increasing throughput
Reducing system footprint, power and cooling overheads
Increasing system operability while reducing IS support overheads
Reducing costs while maintaining adherence to industry standards
Maintaining compatibility with existing infrastructure and workflows
Why Bull for Oil and Gas Industry
Experience in building large HPC systems for government and industry customers in Europe
Bull has implemented several TOP500-class HPC systems for renowned Research Centres as well as many HPC infrastructures of various sizes for renowned companies from the automotive, aerospace, energy and metallurgy sectors in France, Italy and Austria. Delivered by Serviware, the wholly-owned subsidiary of Bull, the current HPC infrastructures in Total EP (Exploration & Production) have evolved significantly since 2004, and are aimed at improving subsurface exploration techniques in the discovery of new hydrocarbon resources. Serviware also recently installed a 17 Tflops Linux cluster at IFP, to be used in the areas of internal combustion, catalysis and simulation of large-scale oil reservoirs.
New technologies and designs e.g. accelerators and GPUs
Bull is the only European IT systems vendor with a mandate to develop leading-edge systems with new technologies, in support of government and industry customers. These include systems with extreme scalability such as the recently-announced collaboration contract to design and build Tera 100, as well as the GENCI project awarded to Bull in April of 2008. The latter includes a new supercomputer of about 300 TFLOPS of which almost 200 TFLOPS of computing power is to be furnished with GPUs, which when deployed will feature radically-improved footprint and FLOPS/watt ratios.
Customer and ISV code adaptation and optimization
In its Competence Centre in the French Alps, Bull has a team of HPC software experts engaged in software development, code testing and benchmarking and long-term research and development of future systems for an increasing number of customer projects. There, customer codes can be examined and optimized for parallel operation on a variety of specific architectures, such as FPGAs, GPUs and multi-core processors.
Technology partnerships and consortia projects
Bull is currently engaging with oil companies’ geoscience data processing and R&D departments to adapt existing codes for use on accelerated architectures, notably GPUs, in collaboration with NVIDIA. Bull offers both a combined approach with the customer’s developers and a turnkey service to optimize code to run on GPGPUs. Bull is also actively consulting with major oil and gas companies and service providers on the desired characteristics of future HPC architectures, scheduled for availability in mid-2009.
Local maintenance and support worldwide
Bull has a well-established presence around the world based on its traditional mainframe technologies as well as on its innovative solutions and services within the public and finance sectors. This presence includes a respected range of customer maintenance and support services available in many countries locally from Bull or its partnership network, including many Africa, Asia and Latin America. Oil companies can therefore be confident that Bull will be able to organize solid maintenance and support for the deployment of Bull-branded and third-party integrated systems in a variety of locations, including remote and challenging places close to oil production sites.
