Performance analysis and early architecture exploration minimizes new technology adoption risks and optimizes the resulting system specification. This paper will demonstrate that early design exploration can validate the specification for performance, power and functionality at an accuracy of 85-90% of the real system. This methodology reduces development time by over 30%, reduces implementation re-spins, and ensures that the required performance, power, functionality, and cost metrics are achieved.Current system architecture exploration approaches take too long, require a lot of custom development, and are not very reusable. Architecture explorations using an analytical approach, such as spreadsheets, have a very limited exploration scope and are prone to errors due to their deterministic and non-graphical nature. Physical prototypes can take a long time to evaluate, and can only be validated for hardware-specific test cases. Graphical models based on Simulink check the correctness of the signal flow and data transfers.Distributed embedded systems must be analyzed for timing, resource usage and power consumption, and spreadsheets typically break down with more than two or three shared resources. Graphical modeling tools can overcome the shared resource concurrency issues, present systems as they are diagrammed, and include built-in statistics and power analysis.
Conceptual Performance Analysis Eliminates Distributed System Design Risks
Sae Technical Papers
SAE World Congress & Exhibition ; 2007
2007-04-16
Aufsatz (Konferenz)
Englisch
Conceptual Performance Analysis Eliminates Distributed System Design Risks
British Library Conference Proceedings | 2007
|Conceptual performance analysis eliminates distributed system design risks
Kraftfahrwesen | 2007
|Holistic design eliminates sewage blockages
Online Contents | 2009
Holistic design eliminates sewage blockages
Online Contents | 2009
Seal-free design eliminates downtime
Online Contents | 2016
|