Platform, United Devices, Fujitsu Unveil Grid Initiatives
Page 1 of 1
Platform Computing and SciTegic have announced Platform Life Sciences Suite, a collaboration the companies say is "the industry's first fully-automated in silico discovery solution."
Platform is combining its workload management software, Platform LSF and Platform JobScheduler, with SciTegic's Pipeline Pilot, a high-throughput data analysis and mining system for drug discovery informatics, to deliver "a comprehensive data pipelining solution that will accelerate time to discovery," the companies said.
"Drug discovery data is being grossly underutilized, due to the combination of increasing volumes and types of data, the dependence upon computational methods, and the inflexibility of legacy systems," Yury Rozenman, Platform's director of life sciences business development, said in a statement. "This solution will accelerate life science research and provide our global customers with a new approach to address this emerging crisis in informatics. SciTegic is pioneering the data pipelining approach, and by integrating with our workload management capabilities, we can ensure that the speed and quality of data analysis is not restricted by the IT system."
By maintaining dependencies among individual execution steps within a pipeline, the companies said the Life Sciences Suite "ensures the automatic execution of each job, while delivering more predictable and reliable results, from R&D to regulatory approval." The solution provides users with visible improvements in the utilization of distributed compute resources, overall performance and scalability, the companies said, and it also provides reusable pipeline protocols, ensuring that intellectual property, processes, and procedures can be easily documented and repeated.
Pipeline Pilot allows disparate data sets to be merged, analyzed, or manipulated and provides support for third party tools like Oracle, ISIS/Host, ISIS/Base, RS3, and Spotfire DecisionSite. Platform Life Sciences Suite integrates with all major Unix and Linux based platforms, including Compaq, HP, IBM, Sun, SGI and Linux.
United Devices, NTT Collaborate On Cell Computing
United Devices and NTT DATA Corp. are cooperating on a large-scale research and development distributed computing project.
NTT DATA is preparing to launch a large-scale trial of cell computing during the first half of the current fiscal year, with participation by a million PCs in Japan with broadband connections, to be followed by commercial service, NTT said.
United Devices has already been involved in a number of large-scale projects using distributed computing technology. Based on the company's existing technology, the joint project will develop a cell computing system geared to broadband users in Japan. The companies will also jointly evaluate the trial results.
Cell computing is a service that NTT DATA said it plans to offer on a commercial basis in the future. The service takes advantage of the unused CPU capacity in PCs in homes and offices that are connected to a broadband network. Using technology capable of harnessing that unused CPU capacity and using it as a virtual supercomputer, the service will provide customers with low-cost CPU power for use in such fields as biotechnology, physics, design, financial engineering, and computer graphics rendering. The PC users who provide their surplus capacity will receive return benefits from this network service. The research for the project is being carried out with the cooperation also of Intel K.K., NTT East Corporation, and SGI Japan.
The companies hope for participation of about one million PCs in Japan with dedicated connections to the Internet, in homes, businesses and organizations, in order to realize massive computing power. That power will be put to use running applications for which there is likely to be popular support, in research fields such as biotechnology, space, and astronomy, NTT said. The trial is expected to last approximately six months starting in the first half of fiscal 2002. NTT said the project is also a contribution to the concept of nationwide broadband service as envisioned in the government's e-Japan Priority Policy Program.
Fujitsu Announces Grid Solution For Sciences
Fujitsu plans to launch in June a new solution for scientific applications called the Grid Solution for the Sciences, which combines the construction of a Grid system environment, education, and operational support.
The solution would assist customers with Grid construction based primarily on Fujitsu's PRIMEPOWER high-performance Unix server (running Solaris), PRIMERGY Intel architecture server (running Linux) and VPP Series supercomputers (running UXP/V).
Working with its research arm, Fujitsu Laboratories Ltd., Fujitsu said it has been a pioneer in Grid development. Fujitsu supplied Grid middleware to the IT-based Laboratory (ITBL) project, which is part of the e-Japan initiative, and, as part of the SuperSINET project, ported the Globus Toolkit Grid middleware to its VPP Series supercomputers installed at the Universities of Kyushu, Kyoto, and Nagoya. In addition, Fujitsu is conducting research on Globus Toolkit portability and other Grid-related technologies in conjunction with the National Institute of Advanced Industrial Science and Technology (AIST), an independent, public sector institution.
Fujitsu is the only Japanese corporate sponsor of the Global Grid Forum, and is participating in the UNICORE project, which is part of EUROGRID, a European Grid research project.
Fujitsu said Grid Solution for the Sciences encompasses a wide range of support services, from software development for the Grid platform to Grid middleware and implementation of applications, and full lifecycle, from design through operation, to help customers construct Grid systems utilizing its high-performance PRIMEPOWER, PRIMERGY and VPP hardware platforms.
DataDirect Networks Pushes Storage For Grids
DataDirect Networks used the recent Tenth NASA Goddard Conference on Mass Storage Systems and Technologies and the Nineteenth IEEE Symposium on Mass Storage Systems at the University of Maryland to showcase its Silicon Storage Appliances for high performance storage area network infrastructures in a modular approach to support terascale and distributed computational grids.
"As distributed grids become more and more prevalent, there are certain foundational requirements for scalability, reliability, performance and manageability," the company said. "DataDirect's storage network appliance technology was designed specifically to enable these larger, extended infrastructures."
"Terascale systems and distributed Grids promise to enable scientific discovery by giving scientists a new, collaborative way to work, and through these systems, teraflops of data that are being collected can be analyzed in ways to allow new insights and knowledge," Robert Woolery, DataDirect's vice president of corporate development, said in a statement.