Specify and use fixed-point and floating-point data types in your design. With fixed-point types, you can specify application-specific word lengths, binary-point scaling, arbitrary slope and bias scaling, and control details such as rounding and overflow modes. Use floating-point simulation to emulate target hardware behavior for denormal floating-point numbers, such as flush-to-zero. You can also simulate limited-precision floating-point data types such as half-precision.
Compare floating-point and fixed-point behavior and analyze tradeoffs in numerical behavior by collecting simulation data and statistics through automatic model-wide instrumentation. Collect range data and use visualizations to optimize your design for efficient hardware resource utilization.