Concurrency Abstraction Watch

 


Home
People
Publications
Multiprocessor Architecture Watch
Concurrency Abstraction Watch
Laboratory

Concurrency Abstraction Watch

The MESCAL project is interested in finding and categorizing design abstractions that seek to help engineers manage concurrency at both the application and architecture levels.

  • An application-level concurrency abstraction is a language programmers can use to declare concurrency intuitively and precisely, without having to specify implementation details. An application abstraction captures concepts that are important for particular application domains without reference to a particular target architecture.
  • An architecture-level concurrency abstraction is a way to express the capabilities of an architectural platform. Its goal is to balance visibility into the architectural features that are necessary to obtain high-performance implementations while hiding details that sap productivity.

We eschew the term programming model. Despite the word "programming" in the name, a traditional programming model is actually an architectural abstraction. The terms architecture abstraction and application abstraction avoid this confusion.

Details coming soon

Send feedback to mescal@gigascale.org
Contact 
©2002-2018 U.C. Regents