Next: Overview Up: High Performance Fortran Language Previous: High Performance Fortran Language

Contents

• Contents
• Overview
  • Goals and Scope of High Performance Fortran
  • Fortran 90 Binding
  • New Features in High Performance Fortran
    • Data Distribution Features
    • Data Parallel Execution Features
    • Extended Intrinsic Functions and Standard Library
    • Extrinsic Procedures
    • Sequence and Storage Association
  • Fortran 90 and Subset HPF
  • Notation
  • HPF-Conforming and Subset-Conforming
  • Journal of Development
    • VIEW Directive
    • Nested WHERE Statements
    • EXECUTE-ON-HOME and LOCAL-ACCESS Directives
    • Elemental Reference of Pure Procedures
    • Parallel I/O
  • Organization of this Document
• High Performance Fortran Terms and Concepts
  • Fortran 90
  • The HPF Model
    • Simple Communication Examples
    • Aggregate Communication Examples
    • Interaction of Communication and Parallelism
  • Syntax of Directives
• Data Alignment and Distribution Directives
  • Model
  • Syntax of Data Alignment and Distribution Directives
  • DISTRIBUTE and REDISTRIBUTE Directives
  • ALIGN and REALIGN Directives
  • DYNAMIC Directive
  • Allocatable Arrays and Pointers
  • PROCESSORS Directive
  • TEMPLATE Directive
  • INHERIT Directive
  • Alignment, Distribution, and Subprogram Interfaces
• Data Parallel Statements and Directives
  • The FORALL Statement
    • General Form of Element Array Assignment
    • Interpretation of Element Array Assignments
    • Examples of the FORALL Statement
    • Scalarization of the FORALL Statement
    • Consequences of the Definition of the FORALL Statement
  • The FORALL Construct
    • General Form of the FORALL Construct
    • Interpretation of the FORALL Construct
    • Examples of the FORALL Construct
    • Scalarization of the FORALL Construct
    • Consequences of the Definition of the FORALL Construct
  • Pure Procedures
    • Pure Procedure Declaration and Interface
      • Pure function definition
      • Pure subroutine definition
      • Pure procedure interfaces
    • Pure Procedure Reference
    • Examples of Pure Procedure Usage
    • Comments on Pure Procedures
  • The INDEPENDENT Directive
    • Examples of INDEPENDENT
    • Visualization of INDEPENDENT Directives
• Intrinsic and Library Procedures
  • Notation
  • System Inquiry Intrinsic Functions
  • Computational Intrinsic Functions
  • Library Procedures
    • Mapping Inquiry Subroutines
    • Bit Manipulation Functions
    • Array Reduction Functions
    • Array Combining Scatter Functions
    • Array Prefix and Suffix Functions
    • Array Sorting Functions
  • Generic Intrinsic and Library Procedures
    • System inquiry intrinsic functions
    • Array location intrinsic functions
    • Mapping inquiry subroutines
    • Bit manipulation functions
    • Array reduction functions
    • Array combining scatter functions
    • Array prefix and suffix functions
    • Array sort functions
  • Specifications of Intrinsic Procedures
    • ILEN(I)
    • MAXLOC(ARRAY, DIM, MASK)
    • MINLOC(ARRAY, DIM, MASK)
    • NUMBER_OF_PROCESSORS(DIM)
    • PROCESSORS_SHAPE()
  • Specifications of Library Procedures
    • ALL_PREFIX(MASK, DIM, SEGMENT, EXCLUSIVE)
    • ALL_SCATTER(MASK,BASE,INDX1, ..., INDXn)
    • ALL_SUFFIX(MASK, DIM, SEGMENT, EXCLUSIVE)
    • ANY_PREFIX(MASK, DIM, SEGMENT, EXCLUSIVE)
    • ANY_SCATTER(MASK,BASE,INDX1, ..., INDXn)
    • ANY_SUFFIX(MASK, DIM, SEGMENT, EXCLUSIVE)
    • COPY_PREFIX(ARRAY, DIM, SEGMENT)
    • COPY_SCATTER(ARRAY,BASE,INDX1, ..., INDXn, MASK)
    • COPY_SUFFIX(ARRAY, DIM, SEGMENT)
    • COUNT_PREFIX(MASK, DIM, SEGMENT, EXCLUSIVE)
    • COUNT_SCATTER(MASK,BASE,INDX1, ..., INDXn)
    • COUNT_SUFFIX(MASK, DIM, SEGMENT, EXCLUSIVE)
    • GRADE_DOWN(ARRAY,DIM)
    • GRADE_UP(ARRAY,DIM)
    • HPF_ALIGNMENT(ALIGNEE, LB, UB, STRIDE, AXIS_MAP, IDENTITY_MAP, DYNAMIC, NCOPIES)
    • HPF_TEMPLATE(ALIGNEE, TEMPLATE_RANK, LB, UB, AXIS_TYPE, AXIS_INFO, NUMBER_ALIGNED, DYNAMIC)
    • HPF_DISTRIBUTION(DISTRIBUTEE, AXIS_TYPE, AXIS_INFO, PROCESSORS_RANK, PROCESSORS_SHAPE)
    • IALL(ARRAY, DIM, MASK)
    • IALL_PREFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • IALL_SCATTER(ARRAY,BASE,INDX1, ..., INDXn, MASK)
    • IALL_SUFFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • IANY(ARRAY, DIM, MASK)
    • IANY_PREFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • IANY_SCATTER(ARRAY,BASE,INDX1, ..., INDXn, MASK)
    • IANY_SUFFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • IPARITY(ARRAY, DIM, MASK)
    • IPARITY_PREFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • IPARITY_SCATTER(ARRAY,BASE,INDX1, ..., INDXn, MASK)
    • IPARITY_SUFFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • LEADZ(I)
    • MAXVAL_PREFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • MAXVAL_SCATTER(ARRAY,BASE,INDX1, ..., INDXn, MASK)
    • MAXVAL_SUFFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • MINVAL_PREFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • MINVAL_SCATTER(ARRAY,BASE,INDX1, ..., INDXn, MASK)
    • MINVAL_SUFFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • PARITY(MASK, DIM)
    • PARITY_PREFIX(MASK, DIM, SEGMENT, EXCLUSIVE)
    • PARITY_SCATTER(MASK,BASE,INDX1, ..., INDXn)
    • PARITY_SUFFIX(MASK, DIM, SEGMENT, EXCLUSIVE)
    • POPCNT(I)
    • POPPAR(I)
    • PRODUCT_PREFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • PRODUCT_SCATTER(ARRAY,BASE,INDX1, ..., INDXn, MASK)
    • PRODUCT_SUFFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • SUM_PREFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
    • SUM_SCATTER(ARRAY,BASE,INDX1, ..., INDXn, MASK)
    • SUM_SUFFIX(ARRAY, DIM, MASK, SEGMENT, EXCLUSIVE)
• Extrinsic Procedures
  • Overview
  • Definition and Invocation of Extrinsic Procedures
  • Requirements on the Called Extrinsic Procedure
• Storage and Sequence Association
  • Storage Association
    • Definitions
    • Examples of Definitions
    • Sequence Directives
    • Storage Association Rules
    • Storage Association Discussion
    • Examples of Storage Association
  • Argument Passing and Sequence Association
    • Sequence Association Rules
    • Discussion of Sequence Association
    • Examples of Sequence Association
• Subset High Performance Fortran
  • Fortran 90 Features in Subset High Performance Fortran
  • Discussion of the Fortran 90 Subset Features
  • HPF Features Not in Subset High Performance Fortran
  • Discussion of the HPF Extension Subset
• Coding Local Routines in HPF and Fortran 90
• Conventions for Local Subprograms
  • Conventions for Calling Local Subprograms
  • Calling Sequence
  • Information Available to the Local Procedure
• Local Routines Written in HPF
  • Restrictions
  • Argument Association
  • HPF Local Routine Library
    • GLOBAL_ALIGNMENT(ARRAY, ...)
    • GLOBAL_DISTRIBUTION(ARRAY, ...)
    • GLOBAL_TEMPLATE(ARRAY, ...)
    • ABSTRACT_TO_PHYSICAL(ARRAY, INDEX, PROC)
    • PHYSICAL_TO_ABSTRACT(ARRAY, PROC, INDEX)
    • LOCAL_TO_GLOBAL(ARRAY, L_INDEX, G_INDEX)
    • GLOBAL_TO_LOCAL(ARRAY, G_INDEX, L_INDEX, LOCAL)
• Local Routines Written in Fortran 90
  • Argument Association
• Example HPF Extrinsic Procedures
• Syntax Rules
• High Performance Fortran Terms and Concepts
  • Syntax of Directives
• Data Alignment and Distribution Directives
  • Syntax of Data Alignment and Distribution Directives
  • DISTRIBUTE and REDISTRIBUTE Directives
  • ALIGN and REALIGN Directives
  • DYNAMIC Directive
  • PROCESSORS Directive
  • TEMPLATE Directive
  • INHERIT Directive
• Data Parallel Statements and Directives
  • The FORALL Statement
  • The FORALL Construct
  • Pure Procedures
  • The INDEPENDENT Directive
• Extrinsic Procedures
  • Definition and Invocation of Extrinsic Procedures
• Storage and Sequence Association
  • Storage Association
• Syntax Cross-reference
• Nonterminal Symbols That Are Defined
• Nonterminal Symbols That Are Not Defined
• Terminal Symbols
• References
• About this document ...

Since its introduction over three decades ago, Fortran has been the language of choice for scientific programming for sequential computers. Exploiting the full capability of modern architectures, however, increasingly requires more information than ordinary FORTRAN 77 or Fortran 90 programs provide. This information applies to such areas as:

• Opportunities for parallel execution;
• Type of available parallelism - MIMD, SIMD, or some combination;
• Allocation of data among individual processor memories; and
• Placement of data within a single processor.

The High Performance Fortran Forum (HPFF) was founded as a coalition of industrial and academic groups working to suggest a set of standard extensions to Fortran to provide the necessary information. Its intent was to develop extensions to Fortran that provide support for high performance programming on a wide variety of machines, including massively parallel SIMD and MIMD systems and vector processors. From its beginning, HPFF included most vendors delivering parallel machines, a number of government laboratories, and many university research groups. Public input was encouraged to the greatest extent possible. The result of this project is this document, intended to be a language specification portable from workstations to massively parallel supercomputers while being able to express the algorithms needed to achieve high performance on specific architectures.

Technical development was carried out by subgroups, and was reviewed by the full committee. Many people served in positions of responsibility:

• Ken Kennedy, Convener and Meeting Chair;
• Charles Koelbel, Executive Director and Head of the FORALL Subgroup;
• Mary Zosel, Head of the Fortran 90 and Storage Association Subgroup;
• Guy Steele, Head of the Data Distribution Subgroup;
• Rob Schreiber, Head of the Intrinsics Subgroup;
• Bob Knighten, Head of the Parallel I/O Subgroup;
• Marc Snir, Head of the Extrinsics Subgroup;
• Joel Williamson and Marina Chen, Heads of the Subroutine Interface Subgroup; and
• David Loveman, Editor.

Geoffrey Fox convened the first HPFF meeting with Ken Kennedy and subsequently led a group to develop benchmarks for HPF. In addition, Clemens-August Thole organized a complementary group in Europe and was instrumental in making this an international effort. Charles Koelbel took notes during every meeting and produced detailed minutes, including summaries of the discussions, that were invaluable to the subgroup heads in preparing successive revisions to the draft proposal. Guy Steele developed macros for a variety of tasks, including formatting BNF grammar, Fortran code and pseudocode, and commentary material; the document would have been much less aesthetically pleasing without his efforts.

Many companies, universities, and other entities supported their employees' attendance at the HPFF meetings, both directly and indirectly. The following organizations were represented at two or more meetings by the following individuals (not including those present at the first HPFF meeting in January of 1992, for which there is no accurate attendee list):

Alliant Computer Systems Corporation David Reese

Amoco Production Company Jerrold Wagener, Rex Page

Applied Parallel Research John Levesque, Rony Sawdayi, Gene Wagenbreth

Archipel Jean-Laurent Philippe

CONVEX Computer Corporation Joel Williamson

Cornell Theory Center David Presberg

Cray Research, Inc. Tom MacDonald, Andy Meltzer

Digital Equipment Corporation David Loveman

Fujitsu America Siamak Hassanzadeh, Ken Muira

Fujitsu Laboratories Hidetoshi Iwashita

GMD-I1.T, Sankt Augustin Clemens-August Thole

Hewlett Packard Maureen Hoffert, Tin-Fook Ngai, Richard Schooler

IBM Alan Adamson, Randy Scarborough, Marc Snir, Kate Stewart

Institute for Computer Applications in Science &Engineering Piyush Mehrotra

Intel Supercomputer Systems Division Bob Knighten

Lahey Computer Lev Dyadkin, Richard Fuhler, Thomas Lahey, Matt Snyder

Lawrence Livermore National Laboratory Mary Zosel

Los Alamos National Laboratory Ralph Brickner, Margaret Simmons

Louisiana State University J. Ramanujam

MasPar Computer Corporation Richard Swift

Meiko, Inc. James Cownie

nCUBE, Inc. Barry Keane, Venkata Konda

Ohio State University P. Sadayappan

Oregon Graduate Institute of Science and Technology Robert Babb II

The Portland Group, Inc. Vince Schuster

Research Institute for Advanced Computer Science Robert Schreiber

Rice University Ken Kennedy, Charles Koelbel

Schlumberger Peter Highnam

Shell Don Heller

State University of New York at Buffalo Min-You Wu

SunPro and Sun Microsystems Prakash Narayan, Douglas Walls

Syracuse University Alok Choudhary, Tom Haupt

TNO-TU Delft Edwin Paalvast, Henk Sips

Thinking Machines Corporation Jim Bailey, Richard Shapiro, Guy Steele

United Technologies Corporation Richard Shapiro

University of Stuttgart Uwe Geuder, Bernhard Woerner, Roland Zink

University of Southampton John Merlin

University of Vienna Barbara Chapman, Hans Zima

Yale University Marina Chen, Aloke Majumdar

Many people contributed sections to the final language specification and HPF Journal of Development, including Alok Choudhary, Geoffrey Fox, Tom Haupt, Maureen Hoffert, Ken Kennedy, Robert Knighten, Charles Koelbel, David Loveman, Piyush Mehrotra, John Merlin, Tin-Fook Ngai, Rex Page, Sanjay Ranka, Robert Schreiber, Richard Shapiro, Marc Snir, Matt Snyder, Guy Steele, Richard Swift, Min-You Wu, and Mary Zosel. Many others contributed shorter passages and examples and corrected errors.

Because public input was encouraged on electronic mailing lists, it is difficult, if not impossible, to identify all of those who contributed to the discussions; the entire mailing list was well over 500 names long. The following list includes some of the active participants in the HPFF process not mentioned above:

The following organizations made the language draft available by anonymous FTP access and/or mail servers: AT&TBell Laboratories, Cornell Theory Center, GMD-I1.T (Sankt Augustin), Oak Ridge National Laboratory, Rice University, Syracuse University, and Thinking Machines Corporation. These outlets were instrumental in distributing the document.

The High Performance Fortran Forum also received a great deal of volunteer effort in nontechnical areas. Theresa Chatman and Ann Redelfs were responsible for most of the meeting planning and organization, including the first HPFF meeting, which drew over 125 people. Shaun Bonton, Rachele Harless, Rhonda Perales, Seryu Patel, and Daniel Swint helped with many logistical details. Danny Powell spent a great deal of time handling the financial details of the project. Without these people, it is unlikely that HPF would have been completed.

HPFF operated on a very tight budget (in reality, it had no budget when the first meeting was announced). The first meeting in Houston was entirely financed from the conferences budget of the Center for Research on Parallel Computation, an NSF Science and Technology Center. DARPA and NSF have supported research at various institutions that have made a significant contribution towards the development of High Performance Fortran. Their sponsored projects at Rice, Syracuse, and Yale Universities were particularly influential in the HPFF process. Support for several European participants was provided by ESPRIT through projects P6643 (PPPE) and P6516 (PREPARE).

R710 R706 R625 R622 R431 R512 R735 R838 R1210 R529 R631 R1221 R1218 R1222 R504 R215 R208 R513 R723 R1209 R1215 R914 R728 R607 R1204 R210 R707 R741 R705 R737 R543 R629 R915 R844 R736 R630 R737 R734 R204 R618 R623 R842 R620 R617 R619 R737 R501 R502 R601 R739 R738 R737

Next: