Real, Shortreal, and Realtime Data Types in Verilog and SystemVerilog

Verilog and SystemVerilog are popular hardware description and verification languages used for designing and testing digital circuits. Among the many data types available in SystemVerilog, both languages support real data types that represent floating-point numbers. This tutorial will cover the three real data types available in both Verilog and SystemVerilog: real, shortreal, and realtime.

Real Data Type in Verilog and SystemVerilog

The real data type is used to represent a 64-bit floating-point number. It is used for simulating analog circuits or for performing mathematical operations that require high precision. The range of values that can be represented by the real data type is from -1.7976931348623157E+308 to 1.7976931348623157E+308, with a precision of approximately 15 digits.

Here is an example of declaring a real variable in Verilog and SystemVerilog:

real myReal = 3.14159;

Shortreal Data Type in Verilog and SystemVerilog

The shortreal data type is used to represent a 32-bit floating-point number. It is commonly used in digital design verification for faster simulation time and lower memory usage. The range of values that can be represented by the shortreal data type is from -3.40282347E+38 to 3.40282347E+38, with a precision of approximately 7 digits.

Here is an example of declaring a shortreal variable in Verilog and SystemVerilog:

shortreal myShortReal = 1.2345;

Realtime Data Type in Verilog and SystemVerilog

The realtime data type is used for modeling time in a simulation. It is a 64-bit floating-point number, just like real, but its value represents simulation time rather than a physical quantity. The realtime data type can be used to model delays and timing constraints in a design.

Here is an example of declaring a realtime variable in Verilog and SystemVerilog:

realtime myTime = $realtime();

Note that numbers declared using the realtime data type cannot be used in edge event control, bit select, or part select, nor can a real or realtime be used as an index in bit select.

Also, when converting a real to an integer, the number is rounded to the nearest integer. If the fractional part of the number is exactly 0.5, it is rounded away from zero. For example, 5.5 is rounded to 6, and -2.5 is rounded to -3.

Conclusion

Real data types are important in Verilog and SystemVerilog for representing floating-point numbers and modeling time in a simulation. Understanding the differences between real, shortreal, and realtime data types can help you choose the appropriate data type for your design needs. By using the right data type, you can optimize your simulations for performance and accuracy.