Index

Abstract processors in HPF

7.3.1 Processors

Actor model

Agglomeration

2.1 Methodical Design, (, )

and granularity: 2.4.1 Increasing Granularity, Surface-to-Volume Effects.
and granularity: 2.4.1 Increasing Granularity, Surface-to-Volume Effects.
design checklist: 2.4.4 Agglomeration Design Checklist
for atmosphere model: Agglomeration.
for floorplan optimization: Agglomeration.
for Fock matrix problem: Communication and Agglomeration.
in data-parallel model: 7.1.3 Design

AIMS performance tool

9.4.7 AIMS, Chapter Notes

Amdahl's law

application to HPF: 7.7.2 Sequential Bottlenecks
definition: 3.2.1 Amdahl's Law, Chapter Notes
definition: 3.2.1 Amdahl's Law, Chapter Notes

Applied Parallel Research

ARPANET

2.3.4 Asynchronous Communication

Asymptotic analysis

limitations of: 3.2.3 Asymptotic Analysis, 3.2.3 Asymptotic Analysis
limitations of: 3.2.3 Asymptotic Analysis, 3.2.3 Asymptotic Analysis
reference: Chapter Notes

Asynchronous communication

in CC++: 5.6 Asynchronous Communication
in FM: 6.5 Asynchronous Communication
in MPI: 8.4 Asynchronous Communication

Asynchronous Transfer Mode

Atmosphere model

basic equations: 2.6.1 Atmosphere Model Background
description: (, )
description: (, )
parallel algorithms: (, )
parallel algorithms: (, )
references: Chapter Notes

BBN Butterfly

Bisection bandwidth

Exercises

Bisection width

Exercises, Chapter Notes

Bitonic mergesort

7.7.2 Sequential Bottlenecks

Bottlenecks in HPF

Branch-and-bound search

description: 2.7.1 Floorplan Background, Chapter Notes
description: 2.7.1 Floorplan Background, Chapter Notes
in MPI: 8.1 The MPI Programming

Breadth-first search

Partition.

Bridge construction problem

definition: 1.3.1 Tasks and Channels
determinism: 1.3.1 Tasks and Channels
in CC++: 5.2 CC++ Introduction
in Fortran M: 6.1 FM Introduction, 6.1 FM Introduction, 6.4.3 Dynamic Channel Structures
in Fortran M: 6.1 FM Introduction, 6.1 FM Introduction, 6.4.3 Dynamic Channel Structures
in Fortran M: 6.1 FM Introduction, 6.1 FM Introduction, 6.4.3 Dynamic Channel Structures
in MPI: 8.2 MPI Basics

Bubblesort

Exercises

Bucketsort

6.5 Asynchronous Communication

Bus-based networks

Bus-based Networks.

Busy waiting strategy

Butterfly

bandwidth competition on: Multistage Interconnection Networks.
description: Replicating Computation.
hypercube formulation: Hypercube Network.

C*

Chapter Notes, 7 High Performance Fortran, Chapter Notes

C++

1.2.2 Other Machine Models, Bus-based Networks.

classes: 5.1.2 Classes
constructor functions: 5.1.2 Classes
default constructors: 5.1.2 Classes
inheritance: 5.1.3 Inheritance, 5.1.3 Inheritance
inheritance: 5.1.3 Inheritance, 5.1.3 Inheritance
member functions: 5.1.2 Classes
overloading: 5.1.1 Strong Typing and
protection: 5.1.2 Classes
virtual functions: 5.1.3 Inheritance

Cache effect

3.6.2 Speedup Anomalies

Cache memory

CC++

Part II: Tools

asynchronous communication: 5.6 Asynchronous Communication
basic abstractions: 5.2 CC++ Introduction
channel communication: 5.5.2 Synchronization
communication costs: 5.10 Performance Issues, 5.10 Performance Issues
communication costs: 5.10 Performance Issues, 5.10 Performance Issues
communication structures: 5.5 Communication
compiler optimization: 5.10 Performance Issues
concurrency: 5.3 Concurrency
library building: 5.11 Case Study: Channel
locality: 5.4 Locality
mapping: (, )
mapping: (, )
mapping: (, )
mapping: (, )
modularity: 5.9 Modularity
modularity: 5.9 Modularity
modularity: 5.9 Modularity
modularity: 5.9 Modularity
nondeterministic interactions: 5.7 Determinism
sequential composition: 5.9 Modularity, 5.9 Modularity
sequential composition: 5.9 Modularity, 5.9 Modularity
synchronization mechanisms: 5.5.2 Synchronization, 5.5.2 Synchronization, 5.5.2 Synchronization, 5.5.3 Mutual Exclusion
synchronization mechanisms: 5.5.2 Synchronization, 5.5.2 Synchronization, 5.5.2 Synchronization, 5.5.3 Mutual Exclusion
synchronization mechanisms: 5.5.2 Synchronization, 5.5.2 Synchronization, 5.5.2 Synchronization, 5.5.3 Mutual Exclusion
synchronization mechanisms: 5.5.2 Synchronization, 5.5.2 Synchronization, 5.5.2 Synchronization, 5.5.3 Mutual Exclusion
threads: 5.2 CC++ Introduction
tutorial: Chapter Notes
unstructured parallelism: 5.3 Concurrency

CHAMMP climate modeling program

Channels

3.8 Input/Output, Chapter Notes

and data dependencies: 1.3.1 Tasks and Channels
connecting outport/inport pairs: 1.3.1 Tasks and Channels
creation in Fortran M: 6.3.1 Creating Channels
dynamic in Fortran M: 6.4.3 Dynamic Channel Structures
for argument passing in Fortran M: 6.7 Argument Passing
in communication: 2.3 Communication
in CSP: Chapter Notes

Checkpointing

CHIMP

1.1.1 Trends in Applications, 2.2.2 Functional Decomposition, 2.6 Case Study: Atmosphere , 9.4.1 Paragraph

Classes in C++

5.1.2 Classes

Climate modeling

in CC++: 5.8.2 Mapping Threads to
in Fortran M: 6.8.3 Submachines
in MPI: 8.8 Case Study: Earth

Clock synchronization

9.3.2 Traces, Chapter Notes

CM Fortran

Collaborative work environments

8.3 Global Operations, 9.4.2 Upshot

Collective communication

Collocation of arrays

7.3.2 Alignment

Combining scatter

7.6.3 HPF Features Not

Communicating Sequential Processes

Surface-to-Volume Effects.

Communication

(, )

and channels: 2.3 Communication
collective: 8.1 The MPI Programming , 8.3 Global Operations
collective: 8.1 The MPI Programming , 8.3 Global Operations
design checklist: 2.3.5 Communication Design Checklist
disadvantages of local: 2.3.2 Global Communication
for atmosphere model: Communication.
for floorplan optimization: Communication.
for Fock matrix problem: Communication and Agglomeration.
in CC++: 5.5 Communication
in data-parallel model: 7.1.3 Design
in Fortran M: 6.3 Communication
in MPI: 8.1 The MPI Programming
synchronous: 6.4.3 Dynamic Channel Structures, 8.6.2 MPI Features Not
synchronous: 6.4.3 Dynamic Channel Structures, 8.6.2 MPI Features Not

Communication costs

Communication Time.

bandwidth competition: 3.7 A Refined Communication
in CC++: 5.10 Performance Issues
in HPF: 7.7.3 Communication Costs
in MPI: 8.7 Performance Issues, 8.7 Performance Issues
in MPI: 8.7 Performance Issues, 8.7 Performance Issues
of unaligned array mapping: 7.7.3 Communication Costs
with cyclic distribution: 7.7.3 Communication Costs

Communication patterns

2.3 Communication

asynchronous: 2.3 Communication, 2.3.4 Asynchronous Communication, 6.5 Asynchronous Communication, 7.6.2 Storage and Sequence , 8.4 Asynchronous Communication
asynchronous: 2.3 Communication, 2.3.4 Asynchronous Communication, 6.5 Asynchronous Communication, 7.6.2 Storage and Sequence , 8.4 Asynchronous Communication
asynchronous: 2.3 Communication, 2.3.4 Asynchronous Communication, 6.5 Asynchronous Communication, 7.6.2 Storage and Sequence , 8.4 Asynchronous Communication
asynchronous: 2.3 Communication, 2.3.4 Asynchronous Communication, 6.5 Asynchronous Communication, 7.6.2 Storage and Sequence , 8.4 Asynchronous Communication
asynchronous: 2.3 Communication, 2.3.4 Asynchronous Communication, 6.5 Asynchronous Communication, 7.6.2 Storage and Sequence , 8.4 Asynchronous Communication
dynamic: 2.3 Communication, 2.3.3 Unstructured and Dynamic
dynamic: 2.3 Communication, 2.3.3 Unstructured and Dynamic
local: 2.3 Communication
many-to-many: 6.4.2 Many-to-Many Communication
many-to-one: 1.4.4 Parameter Study, 6.4.1 Many-to-One Communication
many-to-one: 1.4.4 Parameter Study, 6.4.1 Many-to-One Communication
point-to-point: 8.1 The MPI Programming
static: 2.3 Communication
structured: 2.3 Communication
synchronous: 2.3 Communication, 6.4.3 Dynamic Channel Structures
synchronous: 2.3 Communication, 6.4.3 Dynamic Channel Structures
unstructured: 2.3.3 Unstructured and Dynamic , 6.4 Unstructured Communication
unstructured: 2.3.3 Unstructured and Dynamic , 6.4 Unstructured Communication

Communication time

Communication Time.

Communication/computation ratio

Communicators

seeMPI

Competition for bandwidth

examples: Multistage Interconnection Networks., Multistage Interconnection Networks.
examples: Multistage Interconnection Networks., Multistage Interconnection Networks.
idealized model of: 3.7.1 Competition for Bandwidth
impact: 3.7 A Refined Communication

Compilers

data-parallel: 7.1.3 Design, 7.7.1 HPF Compilation
data-parallel: 7.1.3 Design, 7.7.1 HPF Compilation
for CC++: 5.10 Performance Issues
for Fortran M: 6.10 Performance Issues
for HPF: 7.7.1 HPF Compilation, Chapter Notes
for HPF: 7.7.1 HPF Compilation, Chapter Notes

Composition

concurrent: 4.2 Modularity and Parallel , 4.2.4 Concurrent Composition
concurrent: 4.2 Modularity and Parallel , 4.2.4 Concurrent Composition
definition: 4 Putting Components Together
parallel: 4.2 Modularity and Parallel
sequential: 4.2 Modularity and Parallel , 4.2.2 Sequential Composition
sequential: 4.2 Modularity and Parallel , 4.2.2 Sequential Composition

Compositional C++

seeCC++

Computation time

Computation Time.

Computational chemistry

2.8 Case Study: Computational , Chapter Notes

Computational geometry

1.2.2 Other Machine Models, 3.7.2 Interconnection Networks

Computer architecture

references: Chapter Notes, Chapter Notes, 12 Further Reading
references: Chapter Notes, Chapter Notes, 12 Further Reading
references: Chapter Notes, Chapter Notes, 12 Further Reading
trends: 1.1.4 Summary of Trends

Computer performance improvement

1.1.2 Trends in Computer , 1.1.2 Trends in Computer

Computer trends

1.1.4 Summary of Trends

Computer vision

Chapter Notes, 12 Further Reading

Computer-aided diagnosis

Concert C

Concurrency

explicit vs. implicit: 7.1.1 Concurrency
in CC++: 5.3 Concurrency
in data-parallel programs: 7.1.1 Concurrency
in Fortran M: 6.2 Concurrency
parallel software requirement: 1.1.2 Trends in Computer

Concurrent C

4.2 Modularity and Parallel , 4.2.4 Concurrent Composition

Concurrent composition

benefits: 4.2.4 Concurrent Composition, 4.2.4 Concurrent Composition
benefits: 4.2.4 Concurrent Composition, 4.2.4 Concurrent Composition
cost: 4.2.4 Concurrent Composition
example: 4.2.4 Concurrent Composition
in CC++: 5.8.2 Mapping Threads to
in Fortran M: 6.8.3 Submachines
tuple space example: 4.5 Case Study: Tuple

Concurrent Computation Project

Concurrent data structures

Concurrent logic programming

Conferences in parallel computing

Conformality

definition: 7.1.1 Concurrency
in Fortran M: 6.3.1 Creating Channels
of array sections: 7.2.1 Array Assignment Statement

Constructor functions in C++

5.1.2 Classes

Convolution algorithm

application in image processing: 4.4 Case Study: Convolution
components: 4.4.1 Components
parallel 2-D FFTs: 4.4.1 Components
parallel composition: 4.4.2 Composing Components
sequential composition: 4.4.2 Composing Components

COOL

9.1 Performance Analysis, 9.2.2 Counters

Cosmic Cube

Chapter Notes, Chapter Notes, Chapter Notes

Counters

Cray T3D

1.2.2 Other Machine Models, Chapter Notes

Crossbar switching network

Crossbar Switching Network.

Cycle time trends

1.1.2 Trends in Computer

Cyclic mapping

Cyclic Mappings., Mapping., Mapping., Chapter Notes

in HPF: 7.3.3 Distribution, 7.7.3 Communication Costs, 7.8 Case Study: Gaussian
in HPF: 7.3.3 Distribution, 7.7.3 Communication Costs, 7.8 Case Study: Gaussian
in HPF: 7.3.3 Distribution, 7.7.3 Communication Costs, 7.8 Case Study: Gaussian

Data collection

(, )

basic techniques: 9.1 Performance Analysis
counters: 9.2.2 Counters
process: 9.2.4 Summary of Data
traces: 9.2.3 Traces

Data decomposition

seeDomain decomposition

Data dependency

1.3.2 Other Programming Models, 7 High Performance Fortran

Data distribution

at module boundaries: 4.2.1 Data Distribution
dynamic: 7.6.3 HPF Features Not
in data-parallel languages: 7.1.2 Locality
in HPF: (, )
in HPF: (, )

Data distribution neutrality

benefits: 4.2.1 Data Distribution
example: (, )
example: (, )
in ScaLAPACK: 4.2.2 Sequential Composition
in SPMD libraries: Chapter Notes

Data fitting

3.5.3 Fitting Data to

Data parallelism

and Fortran 90: 7.1.4 Data-Parallel Languages, 7.2.2 Array Intrinsic Functions
and Fortran 90: 7.1.4 Data-Parallel Languages, 7.2.2 Array Intrinsic Functions
and HPF: 7.1.4 Data-Parallel Languages
and modular design: 7.1.3 Design
and task parallelism: Chapter Notes
for irregular problems: Chapter Notes
languages: 7.1.4 Data-Parallel Languages, 9.3.3 Data-Parallel Languages
languages: 7.1.4 Data-Parallel Languages, 9.3.3 Data-Parallel Languages

Data reduction

9.3.1 Profile and Counts, 9.3.2 Traces

Data replication

3.9.3 Shortest-Path Algorithms Summary

Data transformation

9.1 Performance Analysis

Data visualization

9.1 Performance Analysis, 9.3.2 Traces

Data-parallel C

Chapter Notes, Chapter Notes

Data-parallel languages

7.1.4 Data-Parallel Languages, 9.3.3 Data-Parallel Languages

Data-parallel model

1.3.2 Other Programming Models, 7.1.3 Design, 7.1.3 Design, 7.1.3 Design

Databases

Chapter Notes, Chapter Notes, 4.5.1 Application

Deadlock detection

Decision support

Dense matrix algorithms

Depth-first search

Agglomeration.

Design checklists

agglomeration: 2.4.4 Agglomeration Design Checklist
communication: 2.3.5 Communication Design Checklist
mapping: 2.5.3 Mapping Design Checklist
modular design: Design checklist.
partitioning: 2.2.3 Partitioning Design Checklist

Determinism

3.7.1 Competition for Bandwidth

advantages: 1.3.1 Tasks and Channels, Chapter Notes
advantages: 1.3.1 Tasks and Channels, Chapter Notes
in CC++: 5.7 Determinism
in Fortran M: 6.6 Determinism
in MPI: 8.2.2 Determinism

Diagonalization

Exercises, 9.4.2 Upshot

Diameter of network

Dijkstra's algorithm

3.9.2 Dijkstra's Algorithm, 3.9.3 Shortest-Path Algorithms Summary

DINO

Communication and Agglomeration.

DISCO

Distributed computing

1.1.3 Trends in Networking

Distributed data structures

Fock matrix: 2.8 Case Study: Computational
for load balancing: Decentralized Schemes.
implementation: (, )
implementation: (, )
in CC++: 5.12 Case Study: Fock
in Fortran M: 6.11 Case Study: Fock
in MPI: 8.4 Asynchronous Communication
tuple space: 4.5 Case Study: Tuple

Divide-and-conquer

Uncovering Concurrency: Divide

Domain decomposition

2.2 Partitioning, 2.2.1 Domain Decomposition

communication requirements: 2.3 Communication
for atmosphere model: 2.6 Case Study: Atmosphere
for Fock matrix problem: Partition.

Efficiency

3.3.2 Efficiency and Speedup, 3.3.2 Efficiency and Speedup, 3.3.2 Efficiency and Speedup

Embarrassingly parallel problems

1.4.4 Parameter Study

Entertainment industry

1.2.2 Other Machine Models, Chapter Notes

Environmental enquiry;tex2html_html_special_mark_quot;in MPI

8.6.2 MPI Features Not

Ethernet

performance: Communication Time., Ethernet., Multistage Interconnection Networks., Multistage Interconnection Networks.
performance: Communication Time., Ethernet., Multistage Interconnection Networks., Multistage Interconnection Networks.
performance: Communication Time., Ethernet., Multistage Interconnection Networks., Multistage Interconnection Networks.
performance: Communication Time., Ethernet., Multistage Interconnection Networks., Multistage Interconnection Networks.

Event traces

9.1 Performance Analysis, 9.3.2 Traces

Execution profile

3.4.3 Execution Profiles, 3.6 Evaluating Implementations

Execution time

(, )

as performance metric: 3.3 Developing Models
limitations of: 3.3.2 Efficiency and Speedup

Exhaustive search

2.7.1 Floorplan Background

Experimental calibration

3.5.1 Experimental Design, 3.5.1 Experimental Design, 3.5.3 Fitting Data to

Express

Part II: Tools, 8 Message Passing Interface, Chapter Notes, Chapter Notes

Fairness

in CC++: 5.10 Performance Issues
in Fortran M: 6.10 Performance Issues

Fast Fourier transform

4.4 Case Study: Convolution

in convolution: (, )
in convolution: (, )
in HPF: 7.4.2 The INDEPENDENT Directive
performance: Multistage Interconnection Networks.
using hypercube: Chapter Notes

Fine-grained decomposition

2.2 Partitioning

Finite difference algorithm

computation cost: 3.5.3 Fitting Data to
efficiency: 3.3.2 Efficiency and Speedup
execution time: Idle Time.
in CC++: 5.9 Modularity
in Fortran 90: 7.2.2 Array Intrinsic Functions
in Fortran M: 6.9 Modularity
in HPF: 7.3.3 Distribution
in MPI: 8.3.3 Reduction Operations
isoefficiency analysis: 3.4.2 Scalability with Scaled

Finite element method

2.3.3 Unstructured and Dynamic

Fixed problem analysis

3.4.1 Scalability with Fixed

Floorplan optimization problem

description: (, )
description: (, )
parallel algorithms: (, )
parallel algorithms: (, )

Floyd's algorithm

(, ), (, )

Fock matrix problem

algorithms for: Chapter Notes
description: (, )
description: (, )
in CC++: 5.12 Case Study: Fock
in Fortran M: 6.11 Case Study: Fock
in MPI: 8.4 Asynchronous Communication, 8.4 Asynchronous Communication, 8.6.1 Derived Datatypes
in MPI: 8.4 Asynchronous Communication, 8.4 Asynchronous Communication, 8.6.1 Derived Datatypes
in MPI: 8.4 Asynchronous Communication, 8.4 Asynchronous Communication, 8.6.1 Derived Datatypes
performance: 9.4.2 Upshot

Fortran 90

array assignment: 7.2.1 Array Assignment Statement, 7.4 Concurrency
array assignment: 7.2.1 Array Assignment Statement, 7.4 Concurrency
array intrinsics: 7.2.2 Array Intrinsic Functions
as basis for HPF: 7.1.4 Data-Parallel Languages
conformality: 7.1.1 Concurrency, 7.2.1 Array Assignment Statement
conformality: 7.1.1 Concurrency, 7.2.1 Array Assignment Statement
CSHIFT function: 7.2.2 Array Intrinsic Functions
explicit parallelism in: 7.1.1 Concurrency
finite difference problem: 7.2.2 Array Intrinsic Functions
implicit parallelism in: 7.1.1 Concurrency
inquiry functions: 7.6.1 System Inquiry Intrinsic
limitations as data-parallel language: 7.2.2 Array Intrinsic Functions
SIZE function: 7.6.1 System Inquiry Intrinsic
transformational functions: 7.2.2 Array Intrinsic Functions
WHERE: 7.2.1 Array Assignment Statement

Fortran D

Crossbar Switching Network.

Fortran M

Part II: Tools, (, )

and SPMD computations: 6.9 Modularity
argument passing: 6.7 Argument Passing
busy waiting strategy: 6.5 Asynchronous Communication
communication: 6.3 Communication
communication: 6.3 Communication
communication: 6.3 Communication
communication: 6.3 Communication
communication: 6.3 Communication
communication: 6.3 Communication
communication: 6.3 Communication
communication: 6.3 Communication
communication: 6.3 Communication
communication: 6.3 Communication
compiler optimization: 6.10 Performance Issues
concurrency: 6.2 Concurrency
concurrency: 6.2 Concurrency
concurrency: 6.2 Concurrency
concurrency: 6.2 Concurrency
concurrency: 6.2 Concurrency
conformality: 6.3.1 Creating Channels
determinism: 6.6 Determinism, 6.7.1 Copying and Determinism
determinism: 6.6 Determinism, 6.7.1 Copying and Determinism
distribution of data: 6.5 Asynchronous Communication
list of extensions: 6.1 FM Introduction
mapping: (, )
mapping: (, )
mapping: (, )
mapping: (, )
mapping: (, )
message passing: 6.9 Modularity, 6.9 Modularity, 6.9 Modularity
message passing: 6.9 Modularity, 6.9 Modularity, 6.9 Modularity
message passing: 6.9 Modularity, 6.9 Modularity, 6.9 Modularity
modularity: 6.1 FM Introduction
modularity: 6.1 FM Introduction
modularity: 6.1 FM Introduction
modularity: 6.1 FM Introduction
performance analysis: 6.10 Performance Issues
port variables: 6.2.1 Defining Processes
process creation: 6.2.2 Creating Processes
quick reference: 6.12 Summary, 6.12 Summary
quick reference: 6.12 Summary, 6.12 Summary
sequential composition: 6.9 Modularity
tree-structured computation: 6.3.3 Receiving Messages

Fujitsu VPP 500

Functional decomposition

(, )

appropriateness: 2.2.2 Functional Decomposition
communication requirements: 2.3 Communication
complement to domain decomposition: 2.2.2 Functional Decomposition
design complexity reduced by: 2.2.2 Functional Decomposition
for climate model: 2.2.2 Functional Decomposition
for Fock matrix problem: Partition.

Functional programming

Chapter Notes, 12 Further Reading

Gantt chart

9.3.2 Traces, 9.4.1 Paragraph, 9.4.2 Upshot

Gauge performance tool

9.4.4 Gauge, Chapter Notes

Gauss-Seidel update

2.3.1 Local Communication, 2.3.1 Local Communication

Gaussian elimination

7.8 Case Study: Gaussian , 9.3.3 Data-Parallel Languages

Genetic sequences

4.5.1 Application

GIGAswitch

Crossbar Switching Network.

Global communication

2.3.2 Global Communication

Grand Challenge problems

Granularity

2.2 Partitioning

agglomeration used to increase: 2.4 Agglomeration
flexibility related to: 2.2 Partitioning
of modular programs: 4.3 Performance Analysis

Handles in MPI

8.2.1 Language Bindings

Hash tables

4.5.2 Implementation

High Performance Fortran

seeHPF

Histograms

9.3.1 Profile and Counts

HPF

Part II: Tools, (, )

abstract processors: 7.3.1 Processors
advantages: 7.9 Summary
collocation of arrays: 7.3.2 Alignment
compilation: 7.7.1 HPF Compilation
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
data distribution: (, )
extrinsic functions: 7.6.3 HPF Features Not
language specification: Chapter Notes
mapping inquiry functions: 7.6.3 HPF Features Not
modularity: 7.5 Dummy Arguments and
modularity: 7.5 Dummy Arguments and
pure functions: 7.6.3 HPF Features Not
remapping of arguments: Strategy 1: Remap
sequence association: 7.6.2 Storage and Sequence
storage association: 7.6.2 Storage and Sequence
subset (official): 7.1.4 Data-Parallel Languages
system inquiry functions: 7.6.1 System Inquiry Intrinsic

Hypercube algorithms

(, )

all-to-all communication: 11 Hypercube Algorithms
matrix transposition: 11.3 Matrix Transposition
parallel mergesort: 11.4 Mergesort
template for: 11 Hypercube Algorithms
vector broadcast: 11.2 Vector Reduction
vector reduction: 11.2 Vector Reduction, 11.2 Vector Reduction
vector reduction: 11.2 Vector Reduction, 11.2 Vector Reduction

Hypercube network

Hypercube Network.

I/O, parallel

applications requiring: 3.8 Input/Output, Chapter Notes
applications requiring: 3.8 Input/Output, Chapter Notes
performance issues: 3.8 Input/Output, 3.8 Input/Output
performance issues: 3.8 Input/Output, 3.8 Input/Output
two-phase strategy: 3.8 Input/Output, Chapter Notes
two-phase strategy: 3.8 Input/Output, Chapter Notes

IBM RP3

IBM SP

Idle Time., 4.3 Performance Analysis

Idle time

Image processing

Exercises, 4.4 Case Study: Convolution

Immersive virtual environments

9.4.3 Pablo

Incremental parallelization

3.2.1 Amdahl's Law

Information hiding

Ensure that modules

Inheritance in C++

5.1.3 Inheritance

Intel DELTA

3.6.2 Speedup Anomalies, Multistage Interconnection Networks., Multistage Interconnection Networks., Chapter Notes

Intel iPSC

Chapter Notes, Chapter Notes

Intel Paragon

1.2.2 Other Machine Models, Chapter Notes, 9.4.5 ParAide

Intent declarations

6.7.2 Avoiding Copying

Interconnection Networks

seeNetworks

IPS-2 performance tool

3.4.2 Scalability with Scaled , Chapter Notes

Isoefficiency

J machine

2.3.1 Local Communication

Jacobi update

Journals in parallel computing

Kali

10.3.2 The Leapfrog Method, 10.3.2 The Leapfrog Method, 10.3.3 Modified Leapfrog

Latency

3.1 Defining Performance

Leapfrog method

Least-squares fit

3.5.3 Fitting Data to , 3.5.3 Fitting Data to

scaled: 3.5.3 Fitting Data to
simple: 3.5.3 Fitting Data to

Linda

and tuple space: 4.5 Case Study: Tuple , Chapter Notes
and tuple space: 4.5 Case Study: Tuple , Chapter Notes
types of parallelism with: Chapter Notes

Load balancing

cyclic methods: Cyclic Mappings.
dynamic methods: 2.5 Mapping
local methods: 2.5 Mapping, Local Algorithms.
local methods: 2.5 Mapping, Local Algorithms.
manager/worker method: Manager/Worker.
probabilistic methods: 2.5 Mapping, Probabilistic Methods.
probabilistic methods: 2.5 Mapping, Probabilistic Methods.
recursive bisection methods: Recursive Bisection.

Local area network

1.3.2 Other Programming Models

Local communication

definition: 2.3.1 Local Communication
finite difference example: (, )
finite difference example: (, )

Locality

and task abstraction: 1.3.1 Tasks and Channels
definition: 1.2.1 The Multicomputer
in CC++: 5.4 Locality
in data-parallel programs: 7.1.2 Locality, 7.8 Case Study: Gaussian
in data-parallel programs: 7.1.2 Locality, 7.8 Case Study: Gaussian
in multicomputers: 1.2.1 The Multicomputer
in PRAM model: 1.2.2 Other Machine Models

Locks

Machine parameters

Communication Time.

Mapping

2.1 Methodical Design, (, )

design rules: 2.5.3 Mapping Design Checklist
in CC++: 5.8 Mapping, 5.8.2 Mapping Threads to
in CC++: 5.8 Mapping, 5.8.2 Mapping Threads to
in data-parallel model: 7.1.3 Design
in Fortran M: 6.8 Mapping

Mapping independence

MasPar MP

Matrix multiplication

(, )

1-D decomposition: 4.6.1 Parallel Matrix-Matrix Multiplication
2-D decomposition: 4.6.1 Parallel Matrix-Matrix Multiplication, 4.6.1 Parallel Matrix-Matrix Multiplication
2-D decomposition: 4.6.1 Parallel Matrix-Matrix Multiplication, 4.6.1 Parallel Matrix-Matrix Multiplication
and data distribution neutral libraries: 4.6 Case Study: Matrix
communication cost: 4.6.2 Redistribution Costs
communication structure: 4.6.1 Parallel Matrix-Matrix Multiplication
systolic communication: 4.6.3 A Systolic Algorithm

Matrix transpose

seeTranspose

Meiko CS-2

Member functions in C++

5.1.2 Classes

Mentat

Mergesort

(, )

parallel: Compare-Exchange.
parallel algorithms: (, )
parallel algorithms: (, )
performance: Performance
references: Chapter Notes
sequential algorithm: 11.4 Mergesort, 11.4 Mergesort
sequential algorithm: 11.4 Mergesort, 11.4 Mergesort

Mesh networks

Mesh Networks.

Message Passing Interface

seeMPI

Message-passing model

description: Chapter Notes
in HPF: 7.7 Performance Issues
task/channel model comparison: 1.3.2 Other Programming Models

MIMD computers

Modular design

and parallel computing: 1.3 A Parallel Programming , (, )
and parallel computing: 1.3 A Parallel Programming , (, )
and parallel computing: 1.3 A Parallel Programming , (, )
design checklist: Design checklist.
in CC++: 5.9 Modularity
in Fortran M: 6.9 Modularity
in HPF: 7.1.3 Design, 7.5 Dummy Arguments and
in HPF: 7.1.3 Design, 7.5 Dummy Arguments and
in MPI: 8.5 Modularity
in task/channel model: 1.3.1 Tasks and Channels
performance analysis: 4.3 Performance Analysis
principles: (, )
principles: (, )

Monte Carlo methods

MPI

Part II: Tools, (, )

basic functions: 8.2 MPI Basics
C binding: C Language Binding.
collective communication functions: (, )
collective communication functions: (, )
collective communication functions: (, )
collective communication functions: (, )
collective communication functions: (, )
collective communication functions: (, )
collective communication functions: (, )
communicators: 8.5 Modularity, 8.5.1 Creating Communicators
communicators: 8.5 Modularity, 8.5.1 Creating Communicators
communicators: 8.5 Modularity, 8.5.1 Creating Communicators
communicators: 8.5 Modularity, 8.5.1 Creating Communicators
communicators: 8.5 Modularity, 8.5.1 Creating Communicators
communicators: 8.5 Modularity, 8.5.1 Creating Communicators
derived datatypes: 8.6.1 Derived Datatypes, 8.6.1 Derived Datatypes
derived datatypes: 8.6.1 Derived Datatypes, 8.6.1 Derived Datatypes
determinism: 8.2.2 Determinism, 8.2.2 Determinism
determinism: 8.2.2 Determinism, 8.2.2 Determinism
environmental enquiry: 8.6.2 MPI Features Not
Fortran binding: Fortran Language Binding.
handles: 8.2.1 Language Bindings
message tags: 8.2.2 Determinism
modularity: (, )
modularity: (, )
modularity: (, )
modularity: (, )
MPMD model: 8.1 The MPI Programming
performance issues: 8.7 Performance Issues
probe operations: 8.4 Asynchronous Communication
starting a computation: 8.2 MPI Basics

MPI Forum

MPMD model

8.1 The MPI Programming

MPP Apprentice

1.2.1 The Multicomputer, 3.3 Developing Models

Multicomputer model

and locality: 1.2.1 The Multicomputer
early examples: Chapter Notes

Multicomputer Toolbox

1.2.2 Other Machine Models, 1.2.2 Other Machine Models

Multiprocessors

Multistage networks

Multistage Interconnection Networks., Multistage Interconnection Networks.

nCUBE

1.2.2 Other Machine Models, Chapter Notes, Chapter Notes

NESL

Networks

ATM: 1.2.2 Other Machine Models
bus-based: Bus-based Networks.
crossbar switch: Crossbar Switching Network.
Ethernet: Ethernet.
hypercube: Hypercube Network.
LAN: 1.2.2 Other Machine Models
shared memory: Bus-based Networks.
torus: Mesh Networks.
trends in: 1.1.3 Trends in Networking
WAN: 1.2.2 Other Machine Models

Nondeterminism

from random numbers: 3.5.2 Obtaining and Validating
in Fortran M: 6.6 Determinism
in message-passing model: 8.2.2 Determinism
in MPI: 8.2.2 Determinism
in parameter study problem: 1.4.4 Parameter Study, 1.4.4 Parameter Study
in parameter study problem: 1.4.4 Parameter Study, 1.4.4 Parameter Study

Notation

Terminology

Numerical analysis

Object-oriented model

Objective C

3.6.1 Unaccounted-for Overhead

Out-of-core computation

3.8 Input/Output

Overhead anomalies

Overlapping computation and communication

2.4.2 Preserving Flexibility, Idle Time.

Overloading in C++

5.1.1 Strong Typing and

Owner computes rule

7 High Performance Fortran, 7.1.1 Concurrency, 7.8 Case Study: Gaussian

P++ library