自我介绍

Sanshui, Foshan, Guangdong Province, China

我的测试区

Advantages and Disadvantages

Pipelining does not help in all cases. There are several possible disadvantages. An instruction pipeline is said to be fully pipelined if it can accept a new instruction every clock cycle. A pipeline that is not fully pipelined has wait cycles that delay the progress of the pipeline.

并非在所有情况下管线技术都起作用。可能有一些缺点。如果一条指令管线能够在每一个时脉周期(clock cycle)接纳一条新的指令,被称为完整管线化(fully pipelined)。因管线中的指令需要延迟处理而要等待数个时脉周期,被称为非完整管线化。

Advantages of Pipelining:

1. The cycle time of the processor is reduced, thus increasing instruction issue-rate in most cases.

1. 减少了处理器执行指令所需要的时钟周期,在通常情况下增加了指令的输入频率(issue-rate)。

2. Some combinational circuits such as adders or multipliers can be made faster by adding more circuitry. If pipelining is used instead, it can save circuitry vs. a more complex combinational circuit.

2. 一些集成电路(combinational circuits),例如加法器(adders)或者乘法器(multipliers),通过添加更多的环路(circuitry)使其工作得更快。如果以管线化替代,能相对地减少环路。

Disadvantages of Pipelining:

1. A non-pipelined processor executes only a single instruction at a time. This prevents branch delays (in effect, every branch is delayed) and problems with serial instructions being executed concurrently. Consequently the design is simpler and cheaper to manufacture.

1. 非管线化的处理器每次(at a time)只执行一个指令。防止分支延时(事实上,每个分支都会产生延时)和串行指令被并行执行产生的问题。 同时,设计比较简单和较低生产成本。

2. The instruction latency in a non-pipelined processor is slightly lower than in a pipelined equivalent. This is because extra flip flops must be added to the data path of a pipelined processor.

2. 在执行相同的指令时,非管线化处理器的指令传输延迟时间(The instruction latency)比管线化处理器明显较短。这是因为管线化的处理器必须在数据路径(data path)中添加额外正反器(flip-flops)。

3. A non-pipelined processor will have a stable instruction bandwidth. The performance of a pipelined processor is much harder to predict and may vary more widely between different programs.

3. 非管线化处理器有固定指令位宽(a stable instruction bandwidth)。管线化处理器的性能更难以预测,并且不同的程序之间的变化(vary)可能更大。

Execution is as follows: 执行顺序如以下列表所示

Generic pipeline

To the right is a generic pipeline with four stages:

  1. 讀取指令(Fetch)
  2. 指令解碼(Decode)
  3. 執行指令(Execute)
  4. 寫回執行结果(Write-back)

(for lw and sw memory is accessed after execute stage)

The top gray box is the list of instructions waiting to be executed; the bottom gray box is the list of instructions that have been completed; and the middle white box is the pipeline.

上方的大灰色方格是一连串未被执行的指令;下方的大灰色格是已被执行完成的指令;中间的大白色方格是管线。


Time Execution 时序 执行情况 0 Four instructions are awaiting to be executed 四条指令等待执行 1

   * the green instruction is fetched from memory

主存储器(memory)中读取绿色指令

2

   * the green instruction is decoded

绿色指令被解码

   * the purple instruction is fetched from memory

从主存储器中读取紫色指令 3

   * the green instruction is executed (actual operation is performed)

绿色指令被执行(事实上运算已经开始(performed))

   * the purple instruction is decoded

紫色指令被解码

   * the blue instruction is fetched

从主存储器中读取蓝色指令 4

   * the green instruction's results are written back to the register file or memory

绿色指令的运算结果被写回到寄存器(register)或者主存储器

   * the purple instruction is executed

紫色指令被执行

   * the blue instruction is decoded

蓝色指令被解码

   * the red instruction is fetched

从主存储器中读取红色指令

5

   * the green instruction is completed

绿色指令被执行完毕

   * the purple instruction is written back

紫色指令的运算结果被写回到寄存器或者主存储器

   * the blue instruction is executed

蓝色指令被执行

   * the red instruction is decoded

红色指令被解码 6

   * The purple instruction is completed

紫色指令被执行完毕

   * the blue instruction is written back

蓝色指令的运算结果被写回到寄存器或者主存储器

   * the red instruction is executed

红色指令被执行 7

   * the blue instruction is completed

蓝色指令被执行完毕

   * the red instruction is written back

红色指令的运算结果被写回到寄存器或者主存储器 8

   * the red instruction is completed

红色指令被执行完毕 9 All instructions are executed


Bubble 气泡 A bubble in cycle 3 delays execution 一个气泡在编号为3的时脉周期中,指令执行被延迟 Main article: Bubble (computing)

When a "hiccup" in execution occurs, a "bubble" is created in the pipeline in which nothing useful happens. In cycle 2, the fetching of the purple instruction is delayed and the decoding stage in cycle 3 now contains a bubble. Everything "behind" the purple instruction is delayed as well but everything "ahead" of the purple instruction continues with execution.

在指令执行中产生一个“打嗝”(hiccup),在管线中生成一个没有实效的气泡。在编号为2的时脉周期中,紫色指令的读取被延迟,并且在编号为3的时脉周期中解码层也产生了一个气泡。所有在紫色指令之后的指令都被延迟执行,而在其之前已经执行了的指令则不受影响。

Clearly, when compared to the execution above, the bubble yields a total execution time of 8 clock ticks instead of 7.

由于气泡使指令执行延迟了一个时脉周期,完成全部4指令执行共需要8个时脉周期。

Bubbles are like stalls, in which nothing useful will happen for the fetch, decode, execute and writeback. It can be completed with a nop code.