Transaction Level Modeling



While waveforms are great for looking at the results from RTL model simulations, they cannot be used when the modeling abstraction level is raised above where wires or signals are used.

Transaction Level Modeling, or TLM, is an example of this. In TLM, communication takes place via method calls rather than by toggling logic signals. Debugging and analyzing these models tends to involve lots of console printing to understand what is happening. This can be maddening and unproductive, especially for large models or long simulation times.

Scansion's solution to this is to provide the visible representation you need to see what is happening inside your model. It allows you to display a transaction trace and all of the events that occur to this trace over time.

As a viewer, Scansion is language and simulation kernel neutral. Support for transaction traces can be easily added to any simulation environment.

While TLM is a natural usage target for Scansion's transaction trace support, RTL simulations will also benefit from viewing higher level data. Imagine debugging RTL system level simulations by looking at transaction flows and then being able to dive down to the wire-level signals.