October 09, 2006
Last week at the annual meeting of the Society of Exploration Geophysicists in New Orleans, United Devices chief technology officer Jikku Venkat delivered a presentation that described the emerging use of Grid technologies for IT in the oil & gas industry sector. Based on projects United Devices has under way with a number of energy and services companies, Venkat described how he expects Grid technologies to be adopted in the oil & gas space, catching up with other compute-intensive industries like pharmaceuticals, telecommunications, manufacturing, government and financial services.
Information Technology leaders in the
oil & gas sector are a bit like folks from Missouri. Suggest that
Grid technologies can improve their operations, and they skeptically
say, "Show me."
That is a difficult challenge, because
energy companies today exploiting Grid computing to bolster
their clusters or harness non-dedicated resources for high-performance
computing (HPC) view their successes as competitive advantages, and
they are in no hurry to show competitors how to do it. Despite
widespread evaluation of Grid and increasing Grid adoption by energy
and services companies, publicly acknowledged success stories and
project-specific references from the oil patch are rare.
In short, this traditionally conservative industry has been comparatively slow to broadly adopt Grid technologies. That has been the case even though these sophisticated companies live or die by a wide variety of compute-intensive applications that lend themselves extremely well to Grid-enablement, particularly in the disciplines around exploration and production. Instead of being early adopters of Grid, these customers are more likely to either enlarge cluster environments or turn to outside resources for the necessary compute power to run compute-intensive geosciences codes.
But times are changing, and so is the oil & gas industry. A number of energy companies have begun exploring Grid as a strategic technology to help them speed completion of complex computations and transparently optimize use of current computing resources. United Devices has been privileged to work with a number of early adopters who are rapidly realizing the benefits of Grid technologies. As a result, we are confident that the technology is going to see rapid adoption in the oil & gas industry, not just for technical computing, but in the datacenter, as well.
There is strong competition to find replacement reserves economically, and data and compute-intensive technologies like subsurface imaging are on the critical path for these efforts. In addition, newer technologies for reservoir management, such as 4-D seismic imaging for acquisition, processing and interpretation of repeated seismic surveys over time, continue to drive an expanding level of compute power.
Against this challenging engineering and competitive backdrop, Grid computing can play a significant role in optimizing the performance of the industry's increasingly critical IT resources. At United Devices, we see a number of areas where Grid technologies can help: the linking and management of dedicated cluster resources; expansion and management of heterogeneous systems; and the reintroduction of technical computing into a virtual environment within corporate general-purpose datacenters.
Although most oil & gas companies are exploring some form of Grid computing solution, there is no such thing as a typical adopter. The entry point depends on the company, its culture and its existing infrastructure, although certainly PC grids, heterogeneous cluster extensions and Grid-enabled high-performance computing alternatives are among the most typical.
We characterize the oil & gas industry today as being in an early evaluation stage for Grid technology, with adoption for production just starting to take place. From our work with both services and E&P companies, we see a number of solutions gaining significant traction:
The benefits offered by these solutions are:
these scenarios just scratch the surface of Grid technology's potential
for energy companies and, as these projects prove themselves, we expect
further adoption at an increasing rate not only by our customers but
across the whole oil & gas industry. Other than outsourcing to a
third-party capacity provider, the only alternatives companies have are
to either grow their external overflow computing services or continue
to expand their clusters and HPC systems by adding more and more
hardware. Both these options exacerbate the system management
challenges and ultimately will be economically untenable.
Return To The Datacenter
One area ripe for new development is the datacenter, an energy company's central repository of computing power which is today artificially limited in scope to running mainly business applications. For some time, many technical computing operations in oil & gas have shunned the datacenter in favor of dedicated clusters or HPC systems, which they can directly control.
However, in the datacenter of the very near future, virtualized infrastructure managed through Grid technologies will provide new levels of shared compute power while remaining completely transparent to the technical computing user. In short, E&P organizations can tap into the power, efficiency, peak operating capacities, flexibility, scalability and reliability of company datacenter resources without giving up organizational control.
In this new virtualized datacenter, the infrastructure will be managed as a single, shared pool of capacity with little or no application affinity. Applications like reservoir modeling and seismic processing can be readily modified and dynamically bound to the infrastructure as and when needed. Service levels, reliability and availability will be implicitly addressed through automated provisioning and Grid-based system management.
Grid software will automatically detect failures and find replacement devices available and eligible to run applications. It will scale up or scale down resources, automatically finding additional resources when needed or conversely canceling any that are not. Utility computing resources can even be brought into play, tapping such outside suppliers as Hewlett-Packard and IBM.
Implications for geosciences are obvious. The once-shunned datacenters could return as sources of flexible, scalable and comparatively inexpensive additional compute power for technical applications, augmenting today's clusters and dedicated HPC resources. E&P users can maintain control of their applications and pay the datacenter only for capabilities actually used.
Grid computing offers solutions for the breadth of services and oil & gas companies, from independent E&P operators to multinationals. Extending clusters and HPC systems or even tapping into utility computing resources will give independents broad access to large-scale computing resources such as they have never enjoyed. And larger companies and multinationals will be able to harvest new efficiencies and capabilities by adding virtualized services from their existing datacenters to the technical computing mix, cutting down on the need for often wasteful facilities dedicated to specific applications.
Ironically, it is precisely because oil & gas depends so heavily on compute-intensive operations that the industry so often moves cautiously toward newer technologies like Grid computing and Grid-managed data centers. It is understandable. After all, hundreds of millions of dollars often ride on the decisions made based on seismic studies, reservoir models and visualizations.
However, today the energy industry has begun to adopt proven Grid solutions, both to gain more efficient computing power and to provide more robust resources for running ever more complex and numerous technical scenarios. Some of these Grid advances will happen through extended clusters and HPC systems, and others will come as technical computing once again moves toward the power of Grid-managed virtual data centers.
Given wider adoption of Grid technologies by the energy industry, increasingly when IT shops say, "Show me," vendors like United Devices are able to do just that.
** This article originally appeared in SEGwire, HPCwire's exclusive coverage of the Society of Exploration Geophysicists International Exposition.
Jun 19, 2013 |
Ruan Pethiyagoda, Cameron Boehmer, John S. Dvorak, and Tim Sze, trained at San Francisco’s Hack Reactor, an institute designed for intense fast paced learning of programming, put together a program based on the N-Queens algorithm designed by the University of Cambridge’s Martin Richards, and modified it to run in parallel across multiple machines.
Jun 17, 2013 |
With that in mind, Datapipe hopes to establish themselves as a green-savvy HPC cloud provider with their recently announced Stratosphere platform. Datapipe markets Stratosphere as a green HPC cloud service and in doing so partnering with Verne Global and their Icelandic datacenter, which is known for its propensity in green computing.
Jun 12, 2013 |
Cloud computing is gaining ground in utilization by mid-sized institutions who are looking to expand their experimental high performance computing resources. As such, IBM released what they call Redbooks, in part to assist institutions’ movement of high performance computing applications to the cloud.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/02/2012 | AMD | Developers today are just beginning to explore the potential of heterogeneous computing, but the potential for this new paradigm is huge. This brief article reviews how the technology might impact a range of application development areas, including client experiences and cloud-based data management. As platforms like OpenCL continue to evolve, the benefits of heterogeneous computing will become even more accessible. Use this quick article to jump-start your own thinking on heterogeneous computing.