The sequence in which we we define our state variables in solidity smart contracts affects the gas fees we spend while contract deployment as well as while calling functions that make use of those variables.

Solidity stores state variables in 256 bits slots, and if 2(or more) consecutive variable sizes add up to 256 or less, then these variables are stored in same slot. (types like bytes can acquire upto 2 slots)

Let's experiment using following contracts:

contract varSequence_1{
    uint128 var1;
    uint256 var2;
    uint128 var3;
    byte

    function set(uint num1, uint128 num2) external {
        var1= num2;
        var2= num1;
        var3= num2;
    }
}

contract varSequence_2{
    uint128 var1;
    uint128 var3;
    uint256 var2;

    function set(uint num1, uint128 num2) external {
        var1= num2;
        var2= num1;
        var3= num2;
    }
}

Both the contracts are exactly same, just the sequence of defining variables is changed.

Results:

Deployment costs

• varSequence_1

• varSequence_2

There is very minor difference in deployment, let's check the function calls.

Function call costs

• varSequence_1

• varSequence_2

Observation

varSequence_2, which had optimized variable sequence costed almost 25% less than varSequence_1.