본문 바로가기
csapp

5.7.2 Functional Unit Performance

by 정구지개발자 2023. 5. 18.
728x90

 

  • These timings are typical for other proces- sors as well.
  • Each operation is characterized by its latency, meaning the total time required to perform the operation
  • the issue time, meaning the minimum number of clock cycles between two independent operations of the same type
  • the capacity, indicating the number of functional units capable of performing that operation.
  • This short issue time is achieved through the use of pipelining.
  • What this means is that the divider must perform a complete division before it can begin a new one.
  • Having multiple functional units can increase throughput even further.
  •  

 

5.7.3 An Abstract Model of Processor Operation

 

  • As a tool for analyzing the performance of a machine-level program executing on a modern processor, we will use a data-flow representation of programs, a graphical notation showing how the data dependencies between the different operations constrain the order in which they are executed.
728x90

'csapp' 카테고리의 다른 글

From Machine-Level Code to Data-Flow Graphs  (0) 2023.05.20
5.7 Understanding Modern Processors  (0) 2023.05.17
5.3 Program Example  (2) 2023.05.16
5. Optimizing ProgramPerformance  (0) 2023.05.15
Diminishing Returns of Deep Pipelining  (0) 2023.05.08

댓글