Stephan van Beek, Technical Manager at MathWorks in Eindhoven, Netherlands, visited South Africa recently to share his knowledge of how MathWorks tools can be used to develop, test and implement FPGA and SoC designs. He showed how these tools and approaches can increase productivity and decrease development time
Tools shown during the presentation included:
While this event focused on FPGA implementation, MathWorks can support embedded C and C++ code generation. MathWorks and Opti-Num Solutions maintain that MATLAB and Simulink can improve your companies’ efficiency and improve profitability. If you are not convinced, or management requires more substantiated evidence we are happy to discuss a product trial to demonstrate Return on Investment.
During the seminar attendees raised some interesting questions. Here are some questions and answers:
Q: Can you do on target debugging?
A: Yes, FPGA Data Capture in HDL Verifier™ enables you to define signals in FPGA to probe, and automatically generates the components needed to connect your FPGA board to MATLAB or Simulink to analyse signals.
Q: How do I do fixed-point conversion if my model is very large and simulation takes very long?
A: You can used derived range analysis in the Fixed Point Designer. In derived range analysis, the fixed-point tool statically analyses the ranges of individual computations in the model.
Q: Can I do fixed-point conversion if my model has buses?
A: Yes,you can enable automatic data type override.
Q: Is there support for floating point designs?
A: Yes, HDL Coder now supports native floating point.
Q: Does Opti-Num Solutions provide training for HDL Coder?
A: Yes, MathWorks has developed an excellent HDL Coder Course and we have Applications Engineers and Consultants who have presented this course to our clients.
Q: How do I find out more about optimizing HDL Coder generated code?