LLMpediaThe first transparent, open encyclopedia generated by LLMs

Fortran 2003

Generated by GPT-5-mini
Note: This article was automatically generated by a large language model (LLM) from purely parametric knowledge (no retrieval). It may contain inaccuracies or hallucinations. This encyclopedia is part of a research project currently under review.
Article Genealogy
Parent: FORTRAN 77 Hop 4
Expansion Funnel Raw 79 → Dedup 0 → NER 0 → Enqueued 0
1. Extracted79
2. After dedup0 (None)
3. After NER0 ()
4. Enqueued0 ()
Fortran 2003
NameFortran 2003
ParadigmProcedural, object-oriented programming
Year2004
DesignerISO/IEC JTC 1/SC 22
TypingStatic, strong typing
InfluencesFortran 90, Fortran 95
InfluencedFortran 2008

Fortran 2003 Fortran 2003 is a major revision of the Fortran language family published by International Organization for Standardization committees, produced under the auspices of ISO/IEC JTC 1/SC 22 and ratified following work by national bodies including ANSI, DIN, BSI, AFNOR, and JISC. The revision extended prior efforts embodied in Fortran 90 and Fortran 95 and was negotiated among implementers such as Intel Corporation, IBM, Cray Research, and academic contributors from institutions like Massachusetts Institute of Technology, Stanford University, and Lawrence Livermore National Laboratory. The standard influenced subsequent work including Fortran 2008 and discussions at WG5 and other standards fora.

Background and Standardization

The standardization process was driven by ISO working groups and national committees such as IEC, ANSI, and British Standards Institution representatives, with contributions from vendors including Intel, IBM, NAG and research centers such as Los Alamos National Laboratory and Oak Ridge National Laboratory. Early proposals drew on language design research from Edsger W. Dijkstra-influenced procedural theory and practical needs observed at CERN, NCAR, and NASA. The committee deliberations referenced prior standards (Fortran 77, Fortran 90, Fortran 95) and were informed by portability concerns raised by projects like MPI and high-performance initiatives at Argonne National Laboratory and TERAGRID.

Major Language Features

The revision added features to support contemporary scientific computing projects at organizations like DOE laboratories and initiatives such as Human Genome Project-scale simulations. Language additions addressed needs identified by implementers at Intel and Cray and researchers from University of Cambridge and ETH Zurich, including extended procedure constructs used in projects at European Centre for Medium-Range Weather Forecasts and JPL. The new features targeted expression of abstraction and data modeling used in research at MIT, Caltech, and Imperial College London.

Object-Oriented and Type Extensions

Fortran 2003 introduced object-oriented facilities enabling practices adopted by teams at NASA and companies such as Siemens and Boeing, including type extension, inheritance, and polymorphism analogous to mechanisms discussed in Bjarne Stroustrup's work and pedagogy at Carnegie Mellon University. The object model facilitated component-based design used in projects at Lawrence Livermore National Laboratory and influenced hybrid developments at Argonne National Laboratory integrating with C++-based frameworks. Committee papers filed by representatives from IBM and NAG guided method binding, type-bound procedures, and abstract types.

Interoperability and C Binding

A prominent addition was a standardized interoperability mechanism with C (programming language) promoted by contributors from ISO/IEC and implementers such as GNU Project and Intel. The interoperability features were motivated by mixed-language systems developed at Bell Labs and cross-language application integration practiced at Microsoft Research and Oracle Corporation. The C interoperability model enabled linkage to libraries maintained by organizations like NetLib and toolchains used by Red Hat and SUSE in high-performance computing clusters at NERSC and PRACE installations.

Standard Library and Intrinsics

The standard expanded intrinsic procedures and facilities for numerical and control tasks relevant to work at Los Alamos National Laboratory, Sandia National Laboratories, and academic groups at University of Oxford and University of California, Berkeley. Intrinsics and module support were shaped by prior libraries such as BLAS and LAPACK and standards-related discussions involving maintainers from Netlib and researchers associated with Argonne National Laboratory. New intrinsic procedures addressed interoperability, type inquiry, and enhanced array manipulation used in simulation codes at European Space Agency and NOAA.

Compiler Support and Implementations

Compiler vendors including GNU Project (gfortran), Intel Corporation (Intel Fortran), NAG, IBM (XL Fortran), and research compilers at Cray Research and Silverfrost implemented subsets or full support over time, with validation testing undertaken by NIST and conformance suites coordinated by ISO. Implementations were deployed on platforms from x86 clusters operated by CERN to IBM POWER systems used by Fujitsu and supercomputers at Oak Ridge National Laboratory and Argonne National Laboratory.

Adoption, Impact, and Criticism

Adoption occurred in scientific computing centers such as Los Alamos National Laboratory, Lawrence Berkeley National Laboratory, and universities including University of Cambridge and Massachusetts Institute of Technology, and influenced software ecosystems at Netlib and projects like LAMMPS and GROMACS. Critics from academic groups at University of Oxford and industry commentators at ACM noted complexity, partial vendor support, and slow uptake compared with C++ and Python (programming language) in interdisciplinary projects at NASA and European Space Agency, while proponents from Argonne National Laboratory and NAG highlighted benefits for legacy codes and high-performance numerical work.

Category:Programming languages