Performance estimates for applications: An algebraic framework

DOI10.1016/0167-8191(88)90019-1MaRDI QIDQ1114374zbMATH OpenOpenAlexFDO

Authors Dieter Müller-Wichards

Publication date 1988

Published in Parallel Computing (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/0167-8191(88)90019-1

zbMATH Keywords

Amdahl's lawarithmetic operationsbenchmarkingData movementdelaysHockney's lawsordered semigroupparallel-sequential expressionsperformance algebraperformance modellingsemigroup homomorphismsensitivity analysis

Mathematics Subject Classification ID

Ordered semigroups and monoids (06F05) Semigroups in automata theory, linguistics, etc. (20M35) Theory of operating systems (68N25) Theory of software (68N99) Abstract data types; algebraic specification (68Q65)

Review text

The performance of an application is described via a network of its building blocks. If each of the building blocks is assigned a value in the `performance algebra', which turns out to be an ordered semigroup, the total performance can be obtained by applying a semigroup homomorphism. Data movement, arithmetic operations and delays (such as dispatching overhead) and their interactions can be algebraically modelled. The method can be applied to whole applications but also to operations close to the instruction level. Hockney's laws and Amdahl's law can be retrieved as well as the law of the harmonic mean. In addition, a number of inequalities for algebraic constructs are given and the relationship between a `machine independent' and a `machine dependent' application algebra is studied. We formulate an `abstract benchmarking problem' and discuss a number of applications.

Recommendations

• scientific article; zbMATH DE number 4052974
• Process algebra for performance evaluation
• Application research of process algebras on performance evaluation
• Performability analysis: a new algorithm
• Axiomatic-like performance analysis (ALPA)
• Performance analysis of concurrent systems in algebra dtsiPBC

Cited in

(3)

• Computational forces in the linpack benchmark
• Multilevel algebraic approach for performance analysis of parallel algorithms
• Performance coupling: Case studies for improving the performance of scientific applications

This page was built for publication: Performance estimates for applications: An algebraic framework

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1114374)