• the N-bit branch history revisited • branch prediction by branch correlation • tradeoff: using too little context vs trying to use too much • combining predictors • the predicated execution/correlated branches anomaly • The prediction cost matrix
				4