As the reliability of individual
hardware and software components
increases, system failures are more
common. System failures occur in
the interaction of components (hardware,
mechanical and human) rather than
the failure of an individual component.
Most failures are related to flawed
requirements and design constraint-not
implementation errors.
Technology Requirements
The non-deterministic nature of
today's computer systems require
a knowledge of the operating environment,
interrupts handling and incoming
data. A design process requires
planning, analysis and optimization
at the system-level before implementation.
To solve localized and macro scenarios,
it is essential to validate system
specifications against extensive
transaction patterns, applications
and standard benchmark. Current
solutions to evaluate these are
based on rough estimates or creating time-consuming
C-based models. The addition of
a solution to evaluate and model
the performance and architecture
of the system can validate design
assumptions that will otherwise
be identified during development,
testing or customer-use.
 |
Industry
Need
The combining of electronics, hardware,
electro-mechanical and software
within a single product requires
architects and designers to perform
analysis in multiple knowledge areas.
This forces the designer and architects
to learn multiple tools, write non-reusable
C-code or make best-estimates. Moreover,
there is a need for a single platform
to evaluate domains and their dependencies.
The cost of owning and maintaining
multiple tools can overwhelm most
operational budgets and staff.
Most of today’s project teams
are distributed and interpretations
of design specifications are based
on domain knowledge and language
interpretation. The lack of proper
hand-off between teams requires
constant physical interaction or
major untested assumptions creep
in. Hence, even high quality implementation
leads to non-conforming products
that miss market window or requirements
or both.
Customer
Communication
Every company's Web site has reference
block diagrams or graphical descriptions
highlighting the unique capabilities
and differentiated value. Marketing
folks are challenged by the static
nature of these diagrams. Architects
create models of the system that
explore the key values. The combining
of the engineering models with the
marketing needs can make a very
powerful selling tool for companies
either right over the Web or as
a tool for Application Engineers.
VisualSim
Mirabilis Design’s VisualSim
is the first system-level solution
that designs, optimizes and validates
the architecture of systems, sub-systems,
components and embedded software
within a single framework. VisualSim
accomplishes this at a higher level
of abstraction and visualization
through platform-independent technologies.
The product focuses on accelerating
modeling and simulation, IP reuse
and propagation of an executable
simulation as specification. VisualSim
is used for architecture and performance
analysis of hardware, digital ICs
and embedded software. The innovative
methodology provides continuity
from marketing-to-implementation,
thus bridging the gap that is the
major cause of product failure.
The solution includes
an innovative modeling library,
easy applications characterization
and scenario generation to explore
system behavior, performance and power consumption.
Figure 1: VisualSim
Design Methodology |
Flow Description
The standard VisualSim flow starts
at an abtract-level and continues
by the addition of details and refinements.
The final model can be used for
test vector generation, validation
of the implementation and for communication
with customers. The diagram in Figure
1 explains the activity. VisualSim
represents the industry’s
first Top-Down methodology to address
full product exploration and optimization
at the system-level. VisualSim can
combine various abstraction-levels
and analysis needs within a diagram.
