Fortran
A high-level programming language created in 1957, designed for scientific computing and numerical operations. It is known for high performance, backward compatibility, and native array support, and remains a core language for supercomputing and numerical libraries.
Core Facts
Timeline (last 90 days)
OpenAI said its latest model is significantly stronger at understanding legacy codebases such as COBOL, Fortran and even Visual Basic, and can help developers migrate old code without breaking existing functionality.
The KPP language can generate efficient Fortran or C code but cannot integrate into the differentiable programming frameworks that machine learning models rely on.
Languages such as COBOL, Fortran, and Visual Basic still support the critical business systems of banks, governments, and large enterprises.
An AI team announced that its latest model has made significant progress in understanding and transferring low-resource programming languages such as COBOL, Fortran, and Visual Basic.
Low-resource programming languages generally refer to programming languages whose corpus volume in public code repositories, technical forums, documentation and tutorials is far smaller than that of mainstream languages, such as COBOL, Fortran, Ada, Prolog, etc.
Argon and AntiGravity joined forces to autonomously rewrite an old Fortran codebase into C++ and Rust, with the new Rust version running 20% faster.
Since its birth, Fortran has used column-major order memory layout, that is, multidimensional arrays are arranged contiguously by column in memory, with the leftmost index varying fastest.
Coarrays are a native parallel programming mechanism introduced in Fortran 2008, coordinating parallel execution through codimension and built-in procedures such as this_image(), num_images(), and sync all.
DO CONCURRENT is a loop construct introduced in Fortran 2008 and further refined in 2018, used to explicitly declare that there are no order dependencies between loop iterations, allowing the compiler to automatically parallelize or vectorize.
BLAS defines low-level operation interfaces such as vector addition and matrix multiplication, and LAPACK implements higher-level algorithms such as linear equation solving, eigenvalue decomposition, and singular value decomposition on top of it, both of which were originally written in Fortran 77 with publicly available reference implementations.
15 more timeline events
All Facts (20)
Fortran ushered in the era of high-level languages in 1957.
70%UnverifiedOpenAI said its latest model is significantly stronger at understanding legacy codebases such as COBOL, Fortran and even Visual Basic, and can help developers migrate old code without breaking existing functionality.
50%UnverifiedThe KPP language can generate efficient Fortran or C code but cannot integrate into the differentiable programming frameworks that machine learning models rely on.
50%UnverifiedLanguages such as COBOL, Fortran, and Visual Basic still support the critical business systems of banks, governments, and large enterprises.
50%UnverifiedAn AI team announced that its latest model has made significant progress in understanding and transferring low-resource programming languages such as COBOL, Fortran, and Visual Basic.
50%UnverifiedLow-resource programming languages generally refer to programming languages whose corpus volume in public code repositories, technical forums, documentation and tutorials is far smaller than that of mainstream languages, such as COBOL, Fortran, Ada, Prolog, etc.
50%UnverifiedIn many benchmarks, Fortran code is faster than or on par with equivalent C code.
50%UnverifiedFortran was born in 1957 and is a programming language.
50%UnverifiedFortran is fifteen years older than the C language and thirty-four years older than Python.
50%UnverifiedThe standard versions Fortran has gone through include FORTRAN 66, 77, 90, 95, 2003, 2008, 2018, up to the latest 2023 standard.
50%UnverifiedEach standard evolution of Fortran has preserved backward compatibility with old code.
50%UnverifiedThe U.S. Weather Bureau, the European Centre for Medium-Range Weather Forecasts, NASA's simulation systems, and the reference implementations of BLAS and LAPACK all have deep ties to Fortran.
50%UnverifiedSince its birth, Fortran has used column-major order memory layout, that is, multidimensional arrays are arranged contiguously by column in memory, with the leftmost index varying fastest.
50%UnverifiedC/C++ uses row-major memory layout, which is exactly the opposite of Fortran's column-major.
50%UnverifiedCoarrays are a native parallel programming mechanism introduced in Fortran 2008, coordinating parallel execution through codimension and built-in procedures such as this_image(), num_images(), and sync all.
50%UnverifiedDO CONCURRENT is a loop construct introduced in Fortran 2008 and further refined in 2018, used to explicitly declare that there are no order dependencies between loop iterations, allowing the compiler to automatically parallelize or vectorize.
50%UnverifiedBLAS defines low-level operation interfaces such as vector addition and matrix multiplication, and LAPACK implements higher-level algorithms such as linear equation solving, eigenvalue decomposition, and singular value decomposition on top of it, both of which were originally written in Fortran 77 with publicly available reference implementations.
50%UnverifiedThe biggest problem Fortran faces is a talent gap, as most university courses have shifted to Python and C++.
50%UnverifiedThe Fortran community adapts to GPUs and AI accelerator cards through solutions such as OpenMP, OpenACC and CUDA Fortran.
50%UnverifiedEarly SPICE code was written in Fortran or C, which presents a high barrier for modern developers.
50%