Final answer:
A processor would typically need to make two memory references for an indirect-address-mode computation (to fetch the instruction and its operand's address) and for an indirect-address-mode branch instruction (to fetch the instruction and to fetch the branch address).
Step-by-step explanation:
The number of times a processor needs to refer to memory during the fetch and execution of an indirect-address-mode instruction can vary based on the instruction's nature. For a computation requiring a single operand (indirect-address-mode computation), there are typically two memory references: one to fetch the instruction itself and another to fetch the operand's effective address from the memory location given by the instruction.
For a branch instruction utilizing indirect addressing (indirect-address-mode branch), there are generally two memory references as well: one to fetch the instruction and a second to fetch the target address of the branch from memory. However, depending on the specific architecture and the branch condition, there might be additional memory references. For example, evaluating the branch condition itself might require accessing memory if the condition is based on memory-stored data.