October 12, 2010
CERN is an advanced HPC facility located astride the Franco-Swiss border in Geneva and interestingly, is also sharing the gray space on the border between internal grid and cloud computing, according to Sverre Jarp, CERN openlab Chief Technology Officer.
Founded in 1945 as a European center for nuclear research, the nuclear research facility houses vast volumes of experimental data that needs to be stored and processed. This data volume is critical for a variety of CERN affiliates that support new HPC research programs both within the facility and from partner programs at other institutions.
At the core of the CERN HPC information technology infrastructure is the use of their own internal HPC grid that manages the data volumes. Due to the amount of data and complexity of this infrastructure, CERN has created a separate information technology team to handle all aspects of their HPC data volume.
CERN ‘s leadership in the use and development of their own internal HPC grid is exemplary and deserving of a closer look, especially as the program continues to expand, which according to Sverre Jarp, they are building even further on this HPC Grid expertise and “taking steps to move the HPC grid infrastructure into the cloud.”
Data Transformation in CERN’s Internal Grid
While many are already familiar with the CERN facility and the main research objectives, it is less common to encounter information about what takes place with the data collected. CERN developed their own HPC grid specifically for their internal use and application and the methods by which they crunch the vast data sets is based on complex internal application types.
CERN currently uses their own internal HPC grid to collect and store the massive volumes of data generated by the particle accelerators. When the accelerators operate and generate collisions, massive amounts of information must be funneled from the particle detectors to the initial storage units. The collisions take place at 25 nanosecond intervals and it’s the responsibility of the particle detector to pass on this massive, nano-second burst of information.
Data coming from the parcel detector is filtered and sent to the first layer in the CERN HPC grid, the level one filter. The CERN HPC internal grid consists of a massive array of computers for storage and data processing. After residing on the level one filter, the data is then shuttled to the Tier 0 grid layer—a farm made up of thousands of machines. These machines archive the data and send it on to the next level.
Since the data volume is so large, it must be filtered and processed at several levels using a wide range of programs. Tier1 is the next level where the residual data is stored in the HPC grid and shared with major labratories. Finally, this data is transmitted to theTier 2 level, for use at universities and other institutions. These programs that rely on this content for research, analysis and development initiatives.
CERN uses their own internal HPC grid to expand their internal knowledge of HPC grid tools and applications, which they then share with other groups. They have a wealth of information on HPC Grid computing from their own internal use, custom development and refinement of the core technologies.
CERN Stands Alone
CERN has has been leading the initiative in HPC grid computing since its inception. Other institutions have similar programs in concept but their HPC grid centers do generate particle detector data in volumes as massive as CERN. For example, The University of Chicago also has an HPC Grid program. Within their grid program they have added special projects. One of these projects is called ESG Earth System Grid. ESG Earth System Grid integrates grid technology with six other grids. This grid is still being updated and modified and will be the future infrastructure for even more massive data volumes planned with the next operation run of the new ILHC (International Large Hadron Collider).
CERN has expanded their HPC grid technology to include new initiatives and programs that can be shared with the public. Several of these programs are the Openlab, a collaboration between CERN and industrial partners, Open Nebula, an open-source project building the industry standard open source cloud computing tool, Sara Grid program for more grid support and development, Grid Café to share grid technology with the public community and World Wide LHC Grid Computing initiative to build on the existing LHC grid and expand.
Each individual program promotes the concepts and technology already in use by the CERN facility. CERN’s SARA initiative is also doing well. The SARA program tests the use of clusters for HPC computing. SARA project leader, Floris Sluiter received recognition in another HPC Cloud article published last May. Floris states that “we have gotten a very positive evaluation…” of the HPC clustering initiative. This will result in continued support for future endeavors.
During a recent CERN interview with Sverre Jarp, CERN openlab Chief Technology Officer confirmed that their grid now “…treats 20 million jobs per month” . His program is currently working with experts and the latest technology products to expand the HPC grid and “understand cloud infrastructure”. This will be the next area of HPC computing where the cloud will play a significant role. As part of this cloud for HPC grid initiative, his team is looking at “mobile authentication as a standard across all cloud infrastructures”. With mobile authentication, HPC grid technology in the cloud will be secure and robust solution.
What’s Next at the LHC
At openlab, Sverre’s team is currently moving forward with both the HPC Grid and Cloud initiatives “We are leveraging the grid infrastructure to move into the cloud. This is a multi-year initiative…something that won’t happen overnight because of current production initiatives.” However, his team is committed to evolving the program with cloud capabilities and openlab has a pivotal role in leading the technology to move into and HPC cloud. SARA project leader, Floris Sluiter is optimistic. Recent recognition of his SARA HPC clustering program initiatives has resulted in new activities. “We have gotten a go-ahead to turn it into a production service…”.
About the Author
Valery Herrington is CEO of Herrington Technology. She is an enterprise technology leader with Tier 1 Global Technology and a trusted executive advisor and consultant with successful project deployments at over 25 large-scale enterprise organizations. Ms. Herrington specializes in emerging technologies, enterprise architecture, program and project management, and in rapid deployments for the finance, pharmaceutical industries, among others. Ms. Herrington can be found on the web at http://www.herringtontechnology.net
World Wide LHC Grid Computing Project:
World Wide LHC Computing Grid:
Learn About Grid Computing from CERN:
Sara grid program:
University of Chicago:
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
May 10, 2013 |
Australian visual effects company, Animal Logic, is considering a move to the public cloud.
May 10, 2013 |
Program provides cash awards up to $10,000 for the best open-source end-user applications deployed on 100G network.
May 08, 2013 |
For engineers looking to leverage high-performance computing, the accessibility of a cloud-based approach is a powerful draw, but there are costs that may not be readily apparent.
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.