The DSP Algorithm Design solution addresses the exploration and implementation of complex signal processing algorithms influencing the user experience of consumer devices.

DSP Algorithm Exploration and Optimization: Adhoc methodologies using MATLAB/C-modeling are very expensive because they prevent reuse of models and do not scale well across larger design teams. Given the time and cost constraints on the algorithm design team, the specification risk propagates through implementation.

DSP Algorithm Implementation using Fixed Hardware: Traditional paper specification hand-off to the RTL implementation teams increases the likelihood of low-level errors that only show up in real-time measurement during lab integration. One chip re-spin is almost a given.

DSP Algorithm Implementation using Programmable Accelerators: Because of the design challenges associated with it, programmability is seldom added in today's implementation for DSP functions.

CoWare Solution
The CoWare solution allows assembling DSP systems quickly based on a large number of reference models available in C source code and integration with MATLAB. In addition, designers can leverage production-proven, wireless reference libraries for major standards like 3GPP/WCDMA. Using CoWare's optimizing compiled simulation technology, the design space exploration is quick. Floating-point to fixed-point refinement is facilitated by a very clear modeling style and the underlying powerful, yet hidden, C++ infrastructure. Implementation of fixed hardware is supported through production-proven cosimulation and RTL code generation in addition to cosimulation of analog mixed-signal circuits. Programmable accelerators designed with CoWare can be verified through the integration of the generated ISS into the system simulation.