Techniques are described for reducing shared cache memory requests in a multi-threaded microprocessor-based system. One method includes receiving a request for data from a thread, identifying that the request correlates with a pending request associated with a different thread, combining the request with the pending request based on the identifying, and receiving the data after the combining, the receiving being based on the pending request. In some examples, the request may be associated with an address of a cache line in a cache memory.