Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of GraphSearch.
This lecture covers the concept of dynamic scheduling in processor design, aiming to increase parallelism by executing instructions out of order. It discusses breaking the rigidity of basic pipelining, reservation stations, register renaming, and handling exceptions. The lecture also explains in-order and out-of-order completion, WAW and WAR data hazards, and the use of a reorder buffer. It concludes with the implementation of load-store queues and the tangible benefits of dynamic scheduling in terms of instruction-level parallelism.