GOVERNMENT
The 451 Group: Energy Companies Turning to Grid
Analysts at The 451 Group are optimistic about the use of grid computing among oil and gas companies. The companies are turning to grids in response to business pressure to make better use of their resources. 
"Oil and gas companies were early adopters of grids, but their deployments had progressed little beyond those applications for which they were originally procured," said William Fellows, Principal Analyst at The 451 Group. "We now find good reason to be more optimistic about conceded grid use among oil and gas companies. There appears to be more business pressure to make better use of resources, and the application of new techniques to ever-increasing data sets provides a growing opportunity for all kinds of vendors." The 451 Group found that seismic data processing is one area that has dominated high-performance computing (HPC) and grid use across the energy sector. Seismic processing is an overwhelmingly parallel HPC application, and is therefore well suited to running on grids. When it can cost anything from $40m to $100m just to drill a hole in the ground, there are clear business benefits to improving the analysis of seismic data before drilling. The oil and gas industry has another application that can be readily processed on grids - reservoir modeling. This is the analysis of subterranean oil reserves, and it has become ever more important as oil prices rise and oil companies seek to maximize the benefits of reserves. Grid computing enables users to extend conventional data modeling and 3D visualization techniques with other attributes, such as examining the effects of time (for 4D representation) or determining how the application of different operational conditions (more/less pipe pressure, etc.) will affect oil delivery. 451 analysts also found that many oil and gas companies own their own grids or clusters but also have experience in outsourcing to meet some additional capacity requirements. Some companies are awaiting the ability to seamlessly overflow grid queues to an outsourced service provider, which would bill them only for actual use. The issue for vendors is that oil and gas companies can be difficult to support since they typically want to use thousands of CPUs, but only for a very short time. Other sectors mostly purchase longer contracts. "Given the natural resources sector's appetite for outsourced capacity to meet peak workloads and project requirements, ISVs and vendors should enable enterprise IT managers to access these resources as seamlessly as possible. More-flexible license mechanisms, automatic 'overflow' to additional resources and billing for actual use should be the focus for utility-model vendors and ISVs," said Fellows.
"Oil and gas companies were early adopters of grids, but their deployments had progressed little beyond those applications for which they were originally procured," said William Fellows, Principal Analyst at The 451 Group. "We now find good reason to be more optimistic about conceded grid use among oil and gas companies. There appears to be more business pressure to make better use of resources, and the application of new techniques to ever-increasing data sets provides a growing opportunity for all kinds of vendors." The 451 Group found that seismic data processing is one area that has dominated high-performance computing (HPC) and grid use across the energy sector. Seismic processing is an overwhelmingly parallel HPC application, and is therefore well suited to running on grids. When it can cost anything from $40m to $100m just to drill a hole in the ground, there are clear business benefits to improving the analysis of seismic data before drilling. The oil and gas industry has another application that can be readily processed on grids - reservoir modeling. This is the analysis of subterranean oil reserves, and it has become ever more important as oil prices rise and oil companies seek to maximize the benefits of reserves. Grid computing enables users to extend conventional data modeling and 3D visualization techniques with other attributes, such as examining the effects of time (for 4D representation) or determining how the application of different operational conditions (more/less pipe pressure, etc.) will affect oil delivery. 451 analysts also found that many oil and gas companies own their own grids or clusters but also have experience in outsourcing to meet some additional capacity requirements. Some companies are awaiting the ability to seamlessly overflow grid queues to an outsourced service provider, which would bill them only for actual use. The issue for vendors is that oil and gas companies can be difficult to support since they typically want to use thousands of CPUs, but only for a very short time. Other sectors mostly purchase longer contracts. "Given the natural resources sector's appetite for outsourced capacity to meet peak workloads and project requirements, ISVs and vendors should enable enterprise IT managers to access these resources as seamlessly as possible. More-flexible license mechanisms, automatic 'overflow' to additional resources and billing for actual use should be the focus for utility-model vendors and ISVs," said Fellows.