Jump Start e review refreshes your mind about key concepts in using e and Verisity’s Specman Elite to verify your designs. This review reminds you about important e concepts, reiterates helpful tips and methodology you may have overlooked, and ensures everyone is familiar with the elements that are used throughout the course.

Bus functional models are the cornerstone of a proper high-level verification process. When properly designed, they can be highly reusable and configurable. In addition to verifying functions, bus-functional models can verify the timing of output signal timing. This verification can be decoupled from the functional verification to improve simulation performance.

Verification plan development first demonstrates how random, constrained testing can be more productive than a traditional, directed testing approach. Next, the role of functional coverage as a constrained-random approach is explored. Lastly, the verification plan needs to use a different approach than a plan designed for a directed approach. A process for defining a suitable verification plan is defined.

Verification environment details how to implement the appropriate constrainable, random environment structure for the verification strategy you have outlined.

Simulation management explains how the simulations are carried out. Configuration management, directory structure and scripts are discussed. The merits of using compiled e code and using tracing and logging are explained.

Data modeling helps you understand how to model data objects. Object-oriented thinking is presented in the verification context and example data structures are discussed. Implementation techniques for properly translating high-level data types to physical levels are demonstrated. This section concludes with detailed explanations on techniques to improve run-time performance.

Random generation involves more than calling a random number generator function. Random generation must be properly designed to offer the proper level of controllability over all significant parameters, while minimizing the modifications necessary to add or tune constraints. Random generation can also be used to inject errors or protocol violations in bus-functional models.

Self-checking strategies usually involve using scoreboarding techniques. Creating a scoreboard and adding information to optimize performance is explored.

Functional coverage answers the question of how much work is left? This section explores in detail what needs to be verified and when to do it.