Mumbai University > Computer Engineering > sem 8> Parallel and Distributed Systems

Marks: 5M

Data flow architecture uses data driven model in which the program is represented using directed acyclic graph(nodes and edges).

Instructions is represented by a node and the data dependency relationship is represented by the edge between the connected node.

Node can be scheduled for input execution if and if only if data is valid and correct.

Example SISAL, Silage, LISP

Diagram

Advantages

1. It provides very high potential for parallelism.

2. It gives high throughput for a complex computation.

Disadvantages

1. Time is wasted in waiting for unnecessary arguments.

2. High control overhead.

3. Difficult to manipulate data structures.