The Algorithm 10 Architecture Mapping Model (ATAMM) is a new marked graph (a class of Petri net) model for realization and analysis of real-time computing in multicomputer data flow architectures. This paper is concerned with throughput enhancement for periodic execution of large-grain, decision-free algorithms in an homogeneous ATAMM defined data flow architecture. Major applications are expected to be real-time implementation of control, surveillance, and signal processing algorithms where performance is required to be highly predictable. The computing environment is described and performance measures of computing speed and throughput capacity are defined. Bounds on performance measures are established. A technique for improving throughput by insertion of buffers is developed. A software tool to automate the technique in polynomial time is presented.
[1]
K. G. Lockyer.
An introduction to critical path analysis
,
1965
.
[2]
Keshab K. Parhi,et al.
Pipeline interleaving and parallelism in recursive digital filters. II. Pipelined incremental block filtering
,
1989,
IEEE Trans. Acoust. Speech Signal Process..
[3]
Nikitas J. Dimopoulos,et al.
Advanced General Purpose Multicomputer for Space Applications
,
1987,
ICPP.
[4]
George Lee,et al.
A Decomposition Approach for Balancing Large-Scale Acyclic Data Flow Graphs
,
1990,
IEEE Trans. Computers.
[5]
Edward A. Lee,et al.
Static Scheduling of Synchronous Data Flow Programs for Digital Signal Processing
,
1989,
IEEE Transactions on Computers.