Hardware Reference
In-Depth Information
generation of experiments in order to maximize/minimize the metrics following
user guidelines
Less complexity of the flow:
The complexity of the flow is much less in the case
of the automated procedure. The number of steps to be accomplished is much
lower. In this sense, it is interesting to mention that a lower number of steps
represents a lower possibility of errors
Faster convergence times:
The convergence towards the optimum point is
achieved faster with the automated procedure rather than with the semi-automated
procedure
7.2.3.4
Subjective Assessment
In parallel with the objective assessment presented in the precedent paragraph, we
can mention some subjective facts that have been identified by the engineers that
have run the flow in the use case. The main conclusions of this subjective analysis
can be:
MULTICUBE design flow is also a perfect tool to guarantee the coherence and
correctness of the design as we have seen in our study. The analytical capacities
of modeFRONTIER allow to stress the model and to analyze its behavior under
different running conditions allowing investigating if the model behaves as expected
since we can analyze a larger number of operating conditions. As a simulator tool
alone, it adds value to the current simulation methodologies in place.
We can mention that the procedure proposed by MULTICUBE makes the work
of the embedded designer much easier by proposing an excellent front-end tool and
by automating all the processes necessary to update the platform in each simulation
cycle. The intervention of the designer is reduced to a minimum, minimizing also
the possibility of human errors.
MULTICUBE design flow can also benefit from the use of standardized interfaces.
It is possible then to envision the usage of other analysis tools quite easily. In addition
to powerful analysis tools from commercial suites like modeFRONTIER, that have
been proven in other technical domains, we can use other suites developed internally
in the company or to adapt it to future tools.
7.3
Conclusions
This chapter detailed the use of MULTICUBE design flow in the powerline multime-
dia transmission use case for system optimization. It shows that the proposed flow
can save up to a 80% of designer time while achieving better results in terms of per-
formance since much more simulations (up to ten times more possible combinations
than with the semi-automatic design flow) can be run and analyzed. Also, with just
one run we can already identify the optimum for the design parameters. A second or
third iteration would fine tune the design in order to obtain the final parameters.
Search WWH ::
Custom Search