STRUCTURE AND TECHNOLOGY FOR FOLLOWING STORAGE REQUIREMENTS OUT OF Row

摘要

A streaming cache uses multiple dynamically dimensioned tracking queues. Request tracking information is distributed across a variety of tracking queues to selectively return out-of-order storage requests. A dynamically controlled policy assigns pending requests to the tracking queues, providing, for example, out-of-order memory returns in some contexts and/or for a particular traffic and in other contexts and/or out-of-order memory returns for another traffic ready.