KeLP is a C++ framework for implementing portable scientific applications on distributed memory parallel computers including machines like the IBM SP2 and workstation clusters. KeLP is is intended for applications with special needs, in particular, that adapt to data-dependent or hardware dependent conditions at run time. Specific applications include multi-level, multi-resolution, and structured adaptive finite difference methods; multiblock methods; and block structured linear algebra.
KeLP provides run time support for general blocked data decompositions and the management of the underlying irregular communication. The KeLP programmer may customize load balancing activity to the application, and treat interprocessor communication at a very high level. KeLP's ability to support irregular decomposition is also useful on heterogeneous platforms where the processing nodes run at different speeds.
KeLP is in stage 2, and may move to stage 3 in 1 or 2 years.
For more information on KeLP, including a user's guide and applications, please see the URL http://www.cse.ucsd.edu/groups/hpcl/scg/kelp.html.
Last modified: Wed Jan 20 16:40:24 1999