Method and apparatus for memory write performance optimization in architectures with out-of-order read/request-for-ownership response

摘要

A block of data may be transferred to memory through a plurality of write operations, where each write operation is preceded by a protocol request and a protocol response. A plurality of protocol requests issued in a first order may elicit a corresponding plurality of protocol responses in a second order, and the write operations may be performed in yet a third order. Chipsets implementing the data write methods are also described and claimed.