IBM is donating its as yet unreleased Storage Tank software to help international nuclear research centre Cern create the world's largest data grid, supporting its next-generation particle accelerator.
IBM will also initially give Cern 20 terabytes (TB) - equivalent to 20,000GB - of disk storage and six Linux-based servers, and will provide personnel locally in Zurich and at the Storage Tank development centre in Almaden, California.
François Grey, Cern Openlab's development officer, told vnunet.com that a great deal of effort would be needed to build the Large Hadron Collider [LHC] datagrid.
"It is a multi-national project and, being longer-term [it goes live in 2007], it is higher risk because we have to look at what technologies could be viable 4 to 5 years ahead."
He said Storage Tank - now called TotalStorage Common File System (CFS) by IBM - fitted this criterion. IBM is due to release CFS by the end of the year, and Cern is its first major beta site.
Initially the software will handle about one petabyte (PB) of data (equivalent to 1,000TB), the total storage currently at Cern. But this should reach 10PB by the live date, and rise by 10PB a year thereafter.
Storage Tank provides network-based cross-platform file sharing and management. IBM's research centre in Haifa, Israel, will also provide internet small computer system interface storage area network technology for Cern's evaluation.
IBM is the fourth member of Cern's Openlab project for DataGrid applications, the others being Hewlett Packard, Intel and grid software supplier DataSynapse.
Other technologies include an open cluster of 64bit Intel processors. Grey said 64bit should be a commodity by 2007.
Geoengineering on the sea floor near glaciers would form a new ice shelf to prevent melting
Alterations in capillary blood flow can be caused by body position change
Curiosity rover is in 'normal mode' but not transmitting scientific data back to base
NatWest outage comes a day after Barclays' IT systems shut out customers and staff