This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: a. To improve memory performance, b. To prevent access conflicts to different memory locations, c. To prevent data hazards in the pipeline, d. Because FIFO is an efficient storage implementation, or e. Other______________________________________________________________ ANS: (e) to keep memory access in program order. Problem 5: An instruction window has K entries. There are L functional units in our superscalar CPU. Assume an instruction stays inside the instruction window until it completes execution and data is written back to the destination register. An instruction enters into the jth slot of the instruction window. Calculate the number of operand checks the CPU must do for this new instruction. For this problem assume all instructions are R-type. ANS: 3*3*(j-1)...
View Full Document
- Fall '10
- Computer Architecture, Central processing unit, X86, load/store buffer