Associative Arrays in SystemVerilog: A Comprehensive Guide

by Praful Kharade Posted on

In the world of digital design and verification, SystemVerilog provides a variety of array types to handle different data scenarios. One of the most powerful and flexible types is the associative array. Unlike fixed-size or dynamic arrays, associative arrays in SystemVerilog allow indexing using any scalar data type, such as integers, strings, or enumerations, making them highly suitable for sparse data sets and applications where the index values are not sequential.

In this blog, we’ll dive deep into what associative arrays are, why they’re useful, how to use them in SystemVerilog, and compare them to other array types.

What Are Associative Arrays?

Associative arrays are a special kind of array in SystemVerilog that allow you to use arbitrary keys for indexing, rather than being restricted to a continuous range of integers. This means that instead of relying on sequential numerical indices (like 0, 1, 2, etc.), associative arrays let you use a wide range of scalar types, such as integers, strings, or even enums, as the keys.

This flexibility makes them an excellent choice for sparse data sets or cases where the number of elements is unknown at compile time, and indices might not follow a strict order.

Key Characteristics of Associative Arrays

  • Dynamic Sizing: Associative arrays don’t have a predefined size. They grow or shrink dynamically as you add or remove elements.
  • Flexible Indexing: You can use any scalar value as an index, such as integers, strings, or enumerations, making associative arrays extremely flexible.
  • Sparse Data Storage: They are ideal for sparse data storage where only a few keys have valid values and most others are left uninitialized.
  • Efficient Lookups: Associative arrays allow for efficient key-value lookups. If you know the key, you can access the value directly, making operations fast.

Syntax for Associative Arrays in SystemVerilog

The declaration of an associative array is similar to that of other arrays in SystemVerilog, with the difference being in the index type.

Verilog
int myAssociativeArray[string]; // Declare an associative array with string keys

Here, myAssociativeArray is an associative array where the index (key) is of type string, and the values are integers.

Example: Associative Array with Integer Keys

Verilog
module associative_array_example;
  int myAssociativeArray[int]; // Declare an associative array with integer keys

  initial begin
    // Assign values using integer keys
    myAssociativeArray[0] = 10;
    myAssociativeArray[2] = 20;
    myAssociativeArray[100] = 50;

    // Display the values
    $display("myAssociativeArray[0] = %0d", myAssociativeArray[0]);
    $display("myAssociativeArray[2] = %0d", myAssociativeArray[2]);
    $display("myAssociativeArray[100] = %0d", myAssociativeArray[100]);
  end
endmodule

Dynamic Memory Allocation in Associative Arrays

One of the best features of associative arrays in SystemVerilog is that you don’t need to specify the size of the array when you declare it. It grows as you add more elements, making it perfect for scenarios where the size of the dataset isn’t known beforehand.

For example, in the code snippet above, even though the keys 0, 2, and 100 are far apart, SystemVerilog only stores the elements that have been assigned a value. This makes associative arrays efficient in terms of memory usage when working with large, sparse datasets.

Working with Associative Arrays

Iterating Through an Associative Array

Since associative arrays are sparse, you can’t simply use a for loop to iterate through them. Instead, SystemVerilog provides the foreach loop to iterate over all the valid keys in the associative array.

Verilog
module associative_array_example;
  int myAssociativeArray[int]; // Declare an associative array with integer keys

  initial begin
    // Assign values using integer keys
    myAssociativeArray[1] = 10;
    myAssociativeArray[3] = 30;
    myAssociativeArray[7] = 70;

    // Iterate through the associative array
    foreach (myAssociativeArray[i]) begin
      $display("myAssociativeArray[%0d] = %0d", i, myAssociativeArray[i]);
    end
  end
endmodule

In the example above, the foreach loop iterates over all the keys (1, 3, 7) that have been assigned a value and displays their corresponding values.

Checking if an Index Exists

You can check whether a particular index exists in an associative array using the exists() method. This is especially useful when dealing with large, sparse arrays where not all keys will have assigned values.

Verilog
module associative_array_example;
  int myAssociativeArray[int]; // Declare an associative array with integer keys

  initial begin
    myAssociativeArray[5] = 50;

    if (myAssociativeArray.exists(5)) begin
      $display("Key 5 exists with value: %0d", myAssociativeArray[5]);
    end else begin
      $display("Key 5 does not exist");
    end

    if (myAssociativeArray.exists(10)) begin
      $display("Key 10 exists");
    end else begin
      $display("Key 10 does not exist");
    end
  end
endmodule

Advantages of Associative Arrays

  1. Dynamic Sizing: No need to declare a fixed size. Associative arrays automatically expand as elements are added, making them memory-efficient.
  2. Flexible Indexing: Associative arrays allow indexing with more than just integers. You can use strings, enumerations, or even class objects, providing significant flexibility.
  3. Sparse Data Representation: Associative arrays efficiently handle sparse data where only a few indices are populated, without wasting memory.
  4. Efficient Key-Value Lookups: If you know the key, accessing its value is instantaneous, which is ideal for large, sparse datasets.

When to Use Associative Arrays Over Other Array Types?

Dynamic Arrays vs. Associative Arrays:

  • Dynamic arrays are great when you have a continuous range of indices, and you need the array size to grow or shrink during simulation.
  • Associative arrays are better suited for sparse data, where the index values are scattered, or if you need to use non-integer indices.

Fixed-Size Arrays vs. Associative Arrays:

  • Fixed-size arrays are perfect when the size is known in advance and does not change.
  • Associative arrays, on the other hand, provide much more flexibility and are ideal when the index set isn’t known ahead of time or when you need to use more complex data types as keys.

Example: Associative Array with String Keys

Example: Associative Array with String Keys

Verilog
module associative_array_example;
  // Declare an associative array with string keys
  int employeeSalaries[string]; 

  initial begin
    // Assign salaries to employees
    employeeSalaries["Alice"] = 50000;
    employeeSalaries["Bob"] = 60000;
    employeeSalaries["Charlie"] = 55000;

    // Display employee salaries
    foreach (employeeSalaries[name]) begin
      $display("Salary of %s = %0d", name, employeeSalaries[name]);
    end
  end
endmodule

In this example, string keys are used to represent employee names, and their corresponding values represent their salaries. The associative array provides a convenient way to store and access employee information.

Try it on EDA Playground by clicking on the link below

EDA Playground

Conclusion

Associative arrays in SystemVerilog are a versatile and powerful tool, offering dynamic memory management and flexible indexing. They are perfect for handling sparse data and scenarios where the index set is not continuous or even known ahead of time. By understanding how to effectively use associative arrays, you can significantly improve the efficiency and readability of your SystemVerilog code.

If you’re working on large verification environments or dealing with databases of information, associative arrays provide a memory-efficient and fast-access solution.

With their flexibility and ease of use, associative arrays in SystemVerilog are a go-to solution for many verification and design challenges.