SBIR-STTR Award

A c++ virtual shared grid model for architecture independent programming
Award last edited on: 3/8/02

Sponsored Program
SBIR
Awarding Agency
NASA : ARC
Total Award Amount
$49,666
Award Phase
1
Solicitation Topic Code
-----

Principal Investigator
Daniel J Quinlan

Company Information

Front Range Scientific Computations

1390 Claremont Drive
Boulder, CO 80305
   (303) 554-1232
   copper@colorado.edu
   N/A
Location: Single
Congr. District: 02
County: Boulder

Phase I

Contract Number: ----------
Start Date: 00/00/00    Completed: 00/00/00
Phase I year
1992
Phase I Amount
$49,666
This project will investigate the P++ user environment to simplify the development of efficient software for portable use across the widest variety of computer architectures. The major target architectures are distributed memory computers with different kinds of node architectures (vector or superscalar). A simplifying environment for the development of software is needed to take advantage of current and future developments in advanced computational hardware which the P++ environment does by using a standard language, C++, with absolutely no modification of the compiler. Such work is directly related to development of runtime parallel interpretation for FORTRAN 90 D. The advantage of using a standard C++ is that the runtime interpretation of parallelism can be more quickly developed and researched without the construction of a special FORTRAN compiler thereby allowing for the runtime parallel interpretation of FORTRAN to be explored with greater efficiency than would otherwise be possible. Such work also extends the usage of the object-oriented C++ language for development of architecture-independent numerical codes. Such an environment would allow existing C++ language compilers to be used to develop software in the preferred serial environment, and the software could be efficiently run, unchanged, in all target environments. In this way, the investigators introduce an innovative development to permit architecture-independent programming for large-scale scientific applications, which is directly related to existing NASA work on FORTRAN for the parallel environment.

Potential Commercial Applications:
This work will simplify the development of software for advanced computers and will be made generally available as a commercial product through Dyad Software Corp. and Pallas GmbH product lines of tools for the support of scientific programming. Additionally, the runtime environment will be provided a FORTRAN interface that will permit the direct incorporation of the P++ environment into the design of the FORTRAN 90 D compiler.

Phase II

Contract Number: ----------
Start Date: 00/00/00    Completed: 00/00/00
Phase II year
----
Phase II Amount
----