1 / 15

Chapter 1 Microprocessor Architecture and System Concepts

Chapter 1 Microprocessor Architecture and System Concepts. How to improve microprocessor speed. Processor wordlength More bits can be processed internally in parallel Input clock Level of integration Reduce communication overhead between chips Width of external data bus

nicki
Télécharger la présentation

Chapter 1 Microprocessor Architecture and System Concepts

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Chapter 1Microprocessor Architecture and System Concepts

  2. How to improve microprocessor speed • Processor wordlength • More bits can be processed internally in parallel • Input clock • Level of integration • Reduce communication overhead between chips • Width of external data bus • Increase processor-to-memory bandwidth • Architecture advances • Pipelined, superpipelined, superscalar

  3. Microprocessor Performance • Based on the instruction features, microprocessor can be divided to • CICS (complex instruction set computer) • RISC (reduced instruction set computer) • Processor performance • The unit is MIPS (million instructions per second) Processor performance (in MIPS) = F/CPI F: clock rate in MHz CPI: average clock cycle per instruction EX: if operating clock is 40 MHz, the average instruction requires 2 clock cycles, the performance is 20 MIPS The processing time of the program is NI * CPI * C NI: number of instruction in the program C: clock cycle time

  4. System Issues(1) • The modules of a bus-based microprocessor system are divided into two broad classes • Bus master device: • gain control of the bus and are able to initiate data transfers by dividing the address and control lines • Bus slave device • Not capable of controlling the bus • Simple decode the address line and act upon the control signals it receives from the master

  5. System Issues (2): Components • Required main components: • Clock generator • Main memory unit • Coprocessor • MMU (memory management unit) • I/O unit • Communicate with an external device • I/O mapped I/O • Special input and output instruction • Memory mapped I/O • Use memory type instruction to perform the I/O operation

  6. I/O Mapped I/O • 每一個控制器上的暫存器都被給定一個特殊的 I/O 埠。 • Intel 的 IN 跟 OUT 指令可以用來分別讀出或寫入暫存器的值。 CPU AX 控制器 I/O埠 0x68 Outw AX, 0x68 0x15D4 0x15D4

  7. 記憶體 0x0000 0xF000 0xFFFF CPU movw 0xF000, BX movw AX, [BX] AX 0x15D4 控制器暫存器 位址0xF000 0x15D4 Memory Mapped I/O • 記憶體對映 I/O 是將週邊設備的暫存器映對到記憶體位址空間。 • CPU 在存取這些暫存器時,就像是在存取記憶體裡面的值一樣。

  8. 直接記憶體存取 (DMA) • 沒有 DMA 的控制器若想要存取記憶體中的資料,就必須依賴 CPU 的幫忙。 • DMA 控制器的兩個特性 • 可以不經由 CPU 就能存取到記憶體。 • 一次可以處理一整個區塊的資料。

  9. CPU 記憶體 匯流排 無DMA的裝置控制器 有DMA的裝置控制器 I/O裝置 I/O裝置 I/O裝置 DMA存取路徑 無DMA存取路徑 資料可傳遞路徑 DMA存取路徑

  10. How to improve processor performance • Processor performance (in MIPS) = F/CPI • Under constant clock rate  to reduce CPI • How to reduce the CPI • Based on the temporal parallelism • Multiple instructions are simultaneously overlapped in execution using one common hardware • Pipelined, superpipelined • Based on the spatial parallelism • Using separate hardware to execute multiple instructions • superscalar

  11. pipelining

  12. Data dependencies and control hazards along with the execution of conditional-branch instructions

  13. Superpipelined

  14. Superpipelined

  15. Superscalar

More Related