The primary goals of the Agassiz project are to study
the architectural, compiler and system software issues for high-performance and
low-power computer systems. Those systems include both uniprocessors and multiprocessors.
Processor architectural studies are focused on concurrent,
multi-threaded, multi-core architectures that exploit both thread-level and instruction-level
parallelism to achieve high performance with reduced power consupmtion. The targeted systems
span from large-scale parallel machines to application-specific embedded systems.
Compiler effort is focused on a profile-based
approach that supports both medium-grained (loop iteration-level) and fine-grained
(instruction-level) parallelism with speculation, low power and latency hiding schemes.
Research is focused on both static (at compile time) and dynamic (at runtime)
compilation techniques.
System softwawre research currently focuses on operating system support for
multi-core embedded systems, and on supporting virtualization for various
applications.
