Watch FastOS Podcast on Viaway

FastOS is a Department of Energy Office of Science funded program
focused on exploratory work in operating systems and runtimes for petascale
and beyond supercomputers.
Operating and runtime systems provide mechanisms to manage system
hardware and software resources for the efficient execution of
large scale scientific applications. They are essential to the
success of both large scale systems and complex applications. By
the end of this decade petascale computers with thousands of times
more computational power than any in current use will be vital tools
for expanding the frontiers of science and for addressing vital
National priorities. These systems will have tens to hundreds of
thousands of processors, an unprecedented level of complexity, and
will require significant new levels of scalability and fault
management. The overwhelming size and complexity of such systems
poses deep technical challenges that must be overcome to fully
exploit their potential for scientific discovery. Applications require
multiple services from OS/R layers, including: resource management and
scheduling, fault-management (detection, prediction, recovery, and
reconfiguration), configuration management, and file systems access and
management. Current and future large scale parallel systems require that
such services be implemented in a fast and scalable manner so that the
OS/R does not become a performance bottleneck. The current trend in large
scale scientific systems is to leverage operating systems developed for
other areas of computing - operating systems that were not specifically
designed for large scale, parallel computing platforms. Unix, Linux and
other Unix derivatives are the most popular OS's in use for high end
scientific computing, and these all reflect a technological heritage
nearly 30-years old with few fundamental mechanisms to support
parallel systems.

Feedback

FastOS Podcast

You might also like