Quoting from CERN Courier:
The mission of openlab II will continue to be a close collaboration with major trustworthy industrial partners able to demonstrate technological leadership and delivery of viable solutions. In addition, closer collaboration with the European Grid infrastructure project EGEE will be formalized, by taking advantage of the fact that EGEE enters a second phase in May 2006, called EGEE-II. EGEE is currently the largest multiscience Grid-deployment project in the world, with 70 institutional partners in Europe, the US and Russia.
Based on the openlab workshop Industrialising the Grid, held in June 2005, a number of topics were defined as of interest to CERN, the EGEE project and the openlab partners. Two specific themes were singled out as being of particular interest to some of the current openlab partners. These were:
- a Grid interoperability and integration centre (GIC) with close links to EGEE-II
- a platform competence centre (PCC) with the main focus on the PC-based computing hardware and the related software
The PCC will address a range of important fields, such as application optimization and platform virtualization.
Software and hardware optimization is seen as a vital part of the Grid deployment, since the demand for resources by the scientists is very likely to outstrip the available resources, even inside the Grid. Such optimization relies on deep knowledge of the architecture of the entire computing platform. On one hand this covers hardware items, such as processors, memory, buses and input/output channels. On the other hand it covers the ability to use advanced tools, such as profilers, compilers and linkers, specially optimized library functions, etc.
Platform virtualization can allow Grid applications to enjoy a highly secure and standardized environment presented by a “virtual machine hypervisor”, independent of all the hardware and software intricacies. The virtualization concept, initially offered as purely a software solution, will gradually move into hardware, allowing greatly improved performance of both the hypervisors and its “guests”.
Read the whole article at source.