Grids and Computational Resources
Increasingly, high-speed networks feed bits into high-speed computational and data models. A major emphasis of GLORIAD will be to encourage closer work with computer scientists and middleware developers in the US, Russia and China on common, standards-based approaches to grid/software development.
The massive computational and data needs of ATLAS and the CMS experiment at CERN's Large Hadron Collider (LHC) are driving the need for advanced GRID projects. GriPhyN and iVDGL are deeply involved in the development and deployment of Grid infrastructure for these and other frontier experiments in high-energy physics, gravitation wave searches, digital astromony and molecular genomics.
GriPhyN's research program is driven by the need to develop Grid-based architectures that can support such data-intensive experiments and their international science communities. iVDGL is deploying a Grid laboratory where these technologies can be tested at large scale and where advance Grid and networking technologies can be implemented for extended studies by a variety of scientific disciplines. Recently, the Institute for High Energy Physics and Astrophysics (IHEPA) at the University of Florida deployed GRID3, a large-scale, general purpose Grid of approximately 28 sites and 2700 processors that has been running continuously since October 2003.
By early next decade, CERN's LHC experiments together will have accumulated approximately 100 petabytes of data that must be archived, processed, and analyzed by thousands of researchers at institutions across the world. Each experiement will deploy a global data Grid harnessing the entire computing and storage resources of its collaborating scientists into a single resource.
These global Grids will depend crucially on advanced networks to provide througput to allow massive datasets to be moved quickly between CERN and a dozen national computing centers (including those in Russia and China), and the facilities of the participating scientists. The ability to run thousands of simultaneous jobs on the data Grid poses the additional requirement that transfers of multi-terabyte datasets (common for many team-based anlyses) must complete in a few minutes to maintain stability of the overall system. This requirement cannot be met by best-effort networking protocols.
This vision of global data Grids depends fundamentally on the availability of advanced networking and Grid middleware infrastructure in all world regions. GLORIAD provides a critical component of that vision by extending the powerful US-EU networking to Russia and regions in Asia, providing new collaborative links to HEP scientists in those regions.
In the US, the HEP community is a leader in construction and operation of these Grid-based systems. The GLORIAD community includes several advisory board members involved in computational science (leadership of 4 US supercomputing centers), HEP grid developers, the Program Systems Institute of the Russian Academy of Science, PRAGMA, KISTI and the CNIC, which has a large staff devoted to Grid development activities.
- KOREA: Grid Applications and Technology
- With the K*Grid, the Information Technology (IT) can be combined with the emerging technologies such as bio-engineering and nano-technology, as well as traditional engineering fields. In particular, the design of new medicines, practical engineering problems in industry, and the development of a Korean space launcher can be conducted within the framework of the K*Grid.
- KOREA: Research Equipment on Grid
- The Korea Basic Science Institute (KBSI) is the largest provider of state-of-the-art research equipments and large-scale research equipments in Korea. The JEM-ARM 1300S named as "the Morning Star" is the 1st Ultra High Voltage Transmission Electron Microscope (UHV-TEM) installed in 2003. Major functions of the instrument are its atomic resolution, better than 1.2 A (point-to-point), and its precise stage control, capable of tilting more than 60 within 0.1 errors, combination of which is suitable to do 3-D structure analysis of complex materials and the instrument is fully remote controllable without losing its atomic resolution. Specially, Electron tomography offers a powerful tool to obtain three-dimensional (3-D) structural information in biology and materials science.
