Computer Architecture Essentials for IGCSE CCEA Computer Science | 计算机体系结构考点精讲

📚 Computer Architecture Essentials for IGCSE CCEA Computer Science | 计算机体系结构考点精讲

Computer architecture forms the backbone of all computing systems, dictating how a computer’s hardware components are organised and how they interact to execute programs. For the IGCSE CCEA Computer Science specification, a solid understanding of the processor, memory, buses and the fetch‑decode‑execute cycle is essential. This article breaks down every key concept you need to master, with clear explanations, diagrams in words and exam‑ready insights.

计算机体系结构是所有计算系统的基石，它决定了计算机硬件组件如何组织以及如何协同工作以执行程序。对于 IGCSE CCEA 计算机科学课程而言，扎实掌握处理器、存储器、总线以及取指–解码–执行周期至关重要。本文将逐一拆解你需要掌握的核心概念，配以清晰的解释、文字图解和应试要点。

1. What Is Computer Architecture? | 什么是计算机体系结构？

Computer architecture refers to the logical design and functional organisation of a computer system. It specifies how the central processing unit (CPU), memory, input/output devices and the system bus are connected and how data and control signals flow between them. In the IGCSE CCEA course, the Von Neumann architecture is the standard model studied.

计算机体系结构是指计算机系统的逻辑设计和功能组织。它规定了中央处理器（CPU）、存储器、输入/输出设备以及系统总线是如何连接的，以及数据和控制信号如何在它们之间流动。在 IGCSE CCEA 课程中，冯·诺依曼体系结构是需要学习的标准模型。

Understanding architecture allows you to explain why a processor behaves in a certain way, why performance varies and how low‑level programming actually runs on the hardware. It also underpins the fetch‑decode‑execute cycle, which is a guaranteed exam topic.

理解体系结构能让你解释为什么处理器会有某种行为、为什么性能会有所差异以及底层程序到底是如何在硬件上运行的。它也是取指–解码–执行周期这一必考主题的理论基础。

2. The Von Neumann Architecture | 冯·诺依曼架构

The Von Neumann architecture describes a system where program instructions and data share the same memory and are transferred over common buses. Its key components are a control unit (CU), an arithmetic logic unit (ALU), memory (both program and data), input/output devices and the system bus.

冯·诺依曼架构描述了一种系统，其中程序指令与数据共享同一个存储器，并通过公共总线进行传输。其关键组件包括控制单元（CU）、算术逻辑单元（ALU）、存储器（同时存放程序和数据）、输入/输出设备以及系统总线。

Because instructions and data use the same pathways, only one item can be fetched at a time – this is known as the ‘Von Neumann bottleneck’. Despite this limitation, the architecture is simple, cost‑effective and forms the basis of nearly all modern general‑purpose computers.

由于指令和数据共用同一条通路，每次只能读取一个内容——这就是所谓的“冯·诺依曼瓶颈”。尽管存在这一限制，该架构简单、成本效益高，是几乎所有现代通用计算机的基础。

Component 组件	Role in Von Neumann 在冯·诺依曼架构中的作用
Control Unit (CU)	Decodes instructions and sends control signals to coordinate data movement.
控制单元	对指令进行译码，并发出控制信号以协调数据移动。
Arithmetic Logic Unit (ALU)	Performs arithmetic (+, −, ×, ÷) and logic (AND, OR, NOT) operations.
算术逻辑单元	执行算术（+、−、×、÷）和逻辑（AND、OR、NOT）运算。
Memory	Stores both instructions and data in the same read‑write memory (RAM).
存储器	在同一读写存储器（RAM）中同时存放指令和数据。

3. Inside the CPU – Registers, CU and ALU | CPU 内部组件——寄存器、控制单元与 ALU

The CPU contains a set of extremely fast storage locations called registers. Each register has a specific role during the execution of a program. The Program Counter (PC) holds the address of the next instruction to be fetched; the Memory Address Register (MAR) holds any memory address about to be used; the Memory Data Register (MDR) temporarily holds data fetched from or to be written to memory; the Current Instruction Register (CIR) stores the instruction currently being decoded and executed; and the Accumulator (ACC) holds results from the ALU.

CPU 内部包含一组速度极快的存储单元，称为寄存器。每个寄存器在程序执行期间都有特定用途。程序计数器（PC）保存下一条待读取指令的地址；存储器地址寄存器（MAR）保存即将使用的任何存储器地址；存储器数据寄存器（MDR）临时保存从存储器读出或即将写入存储器的数据；当前指令寄存器（CIR）保存正在译码和执行的指令；累加器（ACC）则存放来自 ALU 的运算结果。

The Control Unit orchestrates the whole process. It decodes the binary instruction in the CIR and generates timing and control signals that direct the ALU, registers and buses. The ALU, meanwhile, carries out mathematical and logical calculations as instructed by the CU.

控制单元协调整个过程。它译码 CIR 中的二进制指令，并产生定时和控制信号来指挥 ALU、寄存器和总线。同时，ALU 按照控制单元的指示执行算术和逻辑运算。

PC: points to the next instruction → auto‑increments normally.
PC: 指向下一条指令 → 通常会自增。
MAR: supplies the address for every memory read/write.
MAR: 为每次存储器读/写提供地址。
MDR: acts as a buffer between memory and the CPU.
MDR: 充当内存与 CPU 之间的缓冲。
CIR: splits the instruction into opcode and operand.
CIR: 将指令分割为操作码和操作数。
ACC: intermediate and final arithmetic results live here.
ACC: 存放中间及最终的算术结果。

4. The System Bus – Data, Address and Control Lines | 系统总线——数据总线、地址总线与控制总线

A bus is a set of parallel wires that transfers information between components. The system bus consists of three distinct buses: the data bus, the address bus and the control bus. Each carries a different type of signal and they must work together seamlessly.

总线是一组在组件之间传递信息的并行导线。系统总线由三条独立的总线组成：数据总线、地址总线和控制总线。每条总线传递不同类型的信号，它们必须无缝协作。

The data bus is bidirectional; it carries the actual data or instructions between the processor and memory or I/O devices. Its width (e.g. 8‑bit, 16‑bit, 32‑bit, 64‑bit) determines how much data can be moved in one go. The address bus is unidirectional (from CPU to memory) and carries the address of the memory location being accessed. The number of address lines defines the maximum addressable memory – a 32‑line address bus can address 2³² memory locations. The control bus carries timing and control signals such as read, write, clock and interrupt requests.

数据总线是双向的，在处理器与存储器或 I/O 设备之间传递实际的数据或指令。它的宽度（例如 8 位、16 位、32 位、64 位）决定了一次能移动多少数据。地址总线是单向的（从 CPU 指向存储器），传递被访问的内存单元的地址。地址线的数量决定了可寻址的最大内存空间——一条 32 线的地址总线可以寻址 2³² 个内存单元。控制总线则传递定时和控制信号，如读、写、时钟和中断请求。

Memory capacity = 2^{address lines} × data bus width (bytes)

存储器容量 = 2^地址线数 × 数据总线宽度（字节）

5. The Fetch‑Decode‑Execute Cycle Step by Step | 逐步详解取指–解码–执行周期

Every instruction the CPU processes goes through the same three‑stage cycle. You must be able to describe each stage precisely, using the correct register names.

CPU 处理的每一条指令都经过相同的三阶段周期。你必须能够使用正确的寄存器名称精确描述每个阶段。

Fetch stage: The address in the PC is copied to the MAR. The CU sends a read signal on the control bus. The contents of the addressed memory location travel via the data bus into the MDR. Finally, the PC is incremented (or updated) to point to the next instruction.

取指阶段：PC 中的地址被复制到 MAR。控制单元在控制总线上发出读信号。被寻址的内存单元的内容经数据总线送入 MDR。最后，PC 自增（或更新）以指向下一条指令。

Decode stage: The instruction in the MDR is transferred to the CIR. The CU decodes the binary pattern – the opcode tells the CU what operation is required (e.g. ADD, LOAD) and the operand specifies the data or address involved.

解码阶段：MDR 中的指令被传送到 CIR。控制单元译码该二进制模式——操作码告诉控制单元需要进行什么操作（如 ADD、LOAD），操作数则指定所涉及的数据或地址。

Execute stage: The CU activates the ALU or other components to perform the operation. If the instruction requires reading from memory, the operand address is loaded into the MAR and a read cycle occurs; if it is a write, data moves from ACC to the MDR and then to memory. The result of an arithmetic operation is placed in the ACC. After execution, the cycle repeats, starting with the new PC value.

执行阶段：控制单元激活 ALU 或其他组件以执行操作。如果指令需要从存储器读取数据，操作数地址载入 MAR 并启动读周期；如果是写操作，数据从 ACC 移入 MDR 然后写入存储器。算术运算的结果放入 ACC。执行完成后，周期重复，从新的 PC 值开始。

A diagram in your exam answer should show arrows between PC → MAR, MAR → address bus, MDR ← data bus, MDR → CIR, and then the flow to ALU/ACC.

在考试作答中，你应该画出 PC → MAR、MAR → 地址总线、MDR ← 数据总线、MDR → CIR 以及流向 ALU/ACC 的箭头。

6. Factors Affecting CPU Performance | 影响 CPU 性能的因素

Three hardware characteristics dominate processor performance: clock speed, number of cores and cache memory. CCEA questions often ask you to explain how each one influences execution speed.

三大硬件特征主导了处理器性能：时钟速度、核心数与高速缓存。CCEA 的考题经常要求你解释它们各自如何影响执行速度。

Clock speed, measured in GHz, sets the rhythm of the fetch‑decode‑execute cycle. Each cycle advances the processor by one tick; a 3 GHz clock means 3 × 10⁹ cycles per second. Higher clock speeds allow more instructions to be processed per unit time, but they also generate more heat and may be limited by the speed of other components.

时钟速度以 GHz 为单位，它设定了取指–解码–执行周期的节拍。每个周期让处理器前进一个节拍；3 GHz 时钟意味着每秒 3×10⁹ 个周期。更高的时钟速度能在单位时间内处理更多指令，但同时也会产生更多热量，并可能受到其他组件速度的限制。

Number of cores: A dual‑core or quad‑core processor contains multiple complete CPUs on one chip. They can run multiple instructions truly simultaneously (parallel processing), provided the software is written to distribute tasks. More cores do not always give a simple doubling of speed; there is overhead in coordinating tasks.

核心数量：双核或四核处理器在一个芯片上包含多个完整的 CPU。只要软件经过编写以分配任务，它们就能真正同时执行多条指令（并行处理）。更多的核心并不总是让速度简单翻倍；协调任务会带来额外开销。

Cache memory: Cache is a small, extremely fast memory located on or very close to the CPU. It holds frequently used instructions and data so the processor can access them without waiting for slower RAM. L1 cache is the fastest but smallest, L2 is larger but slightly slower, and L3 cache is shared among cores. A larger cache generally improves performance because the CPU spends less time waiting.

高速缓存：高速缓存是位于 CPU 内部或非常靠近 CPU 的小型、极快存储器。它保存常用的指令和数据，以便处理器无需等待较慢的 RAM 即可访问它们。L1 缓存最快但最小，L2 更大但稍慢，L3 缓存在多个核心之间共享。更大的缓存通常能提升性能，因为 CPU 等待的时间减少了。

Execution time ≈ (Instructions × CPI) / Clock rate

执行时间 ≈ （指令数 × 每指令周期数）/ 时钟频率

7. Memory Hierarchy and the Role of Storage | 存储层次结构与存储器的作用

Computers use a hierarchy of memory types to balance speed and cost. From fastest and most expensive to slowest and cheapest: registers, cache (L1, L2, L3), RAM, and secondary storage such as HDDs, SSDs and optical disks. Data that is accessed frequently moves up the hierarchy; rarely used data stays lower down.

计算机利用存储器类型的层次结构来平衡速度与成本。从最快最贵到最慢最便宜依次为：寄存器、高速缓存（L1, L2, L3）、RAM，以及二级存储器，如硬盘驱动器、固态硬盘和光盘。频繁访问的数据会上移到层次结构的顶端，较少使用的数据则停留在较低的层次。

RAM (Random Access Memory) is volatile main memory that holds the operating system, applications and data currently in use. It connects directly to the processor via the system bus. ROM (Read Only Memory) is non‑volatile and stores the BIOS or boot firmware; its contents survive a power cycle.

RAM（随机存取存储器）是易失性的主存储器，保存着当前正在使用的操作系统、应用程序和数据。它通过系统总线直接连接到处理器。ROM（只读存储器）是非易失性的，存储着 BIOS 或引导固件；其内容在断电后依然保留。

Secondary storage is non‑volatile and holds data permanently. Magnetic storage (HDD) uses spinning platters; optical storage (CD, DVD, Blu‑ray) uses lasers; solid‑state storage (SSD, USB flash) uses NAND flash chips. SSDs are much faster and more shock‑resistant than HDDs, but typically cost more per gigabyte.

二级存储器是非易失性的，可永久保存数据。磁性存储器（HDD）使用旋转盘片；光学存储器（CD、DVD、蓝光）使用激光；固态存储器（SSD、USB 闪存）使用 NAND 闪存芯片。SSD 比 HDD 快得多且更抗震，但每吉字节成本通常更高。

Memory Type 存储类型	Volatile? 易失性	Typical Speed	Purpose 用途
Registers 寄存器	Yes	Fastest (sub‑ns)	Immediate data for ALU/CU
Cache 高速缓存	Yes	~1‑10 ns	Frequent instructions/data
RAM	Yes	~10‑100 ns	Running programs, OS
SSD / HDD 固态硬盘/机械硬盘	No	Milliseconds	Long‑term file storage 长期文件存储

8. Embedded Systems – A Specialised Architecture | 嵌入式系统——一种专用架构

An embedded system is a microprocessor‑based computer system designed to perform a dedicated function within a larger mechanical or electrical system. Unlike a general‑purpose desktop PC, an embedded system runs firmware stored in ROM or flash memory and often has very limited user interaction.

嵌入式系统是一种基于微处理器的计算机系统，设计用于在更大的机械或电气系统中执行专用功能。与通用台式电脑不同，嵌入式系统运行存储在 ROM 或闪存中的固件，并且通常具有非常有限的用户交互。

Common examples include washing machine controllers, digital watches, car engine management units, traffic lights and smart thermostats. These devices prioritise low power consumption, small physical size, real‑time response and high reliability. They are usually cheaper because they contain only the necessary hardware – no hard drive, no keyboard, a tailored set of I/O ports.

常见的例子包括洗衣机控制器、电子手表、汽车发动机管理单元、交通信号灯和智能恒温器。这些设备优先考虑低功耗、小尺寸、实时响应和高可靠性。它们通常更便宜，因为只包含必要的硬件——没有硬盘、没有键盘，只有一组定制的 I/O 端口。

In the CCEA exam, you may be asked to compare an embedded processor with a standard desktop CPU, highlighting differences in purpose, memory, operating system (often real‑time OS or no OS) and upgradability. Embedded systems are typically not user‑programmable once deployed.

在 CCEA 考试中，你可能需要比较嵌入式处理器与标准台式机 CPU，突出它们在用途、存储器、操作系统（通常是实时操作系统或无操作系统）和可升级性方面的差异。嵌入式系统在部署后通常不能再由用户编程。

9. Common Exam Pitfalls and How to Avoid Them | 常见考试陷阱及应对策略

One of the biggest mistakes is confusing registers – for instance mixing up MAR (address) with MDR (data). Remember that the MAR always holds an address, and the MDR holds the actual value. The PC holds an address too, but it is specifically the next instruction’s address.

最大的错误之一就是混淆寄存器——例如将 MAR（地址）与 MDR（数据）搞混。请记住，MAR 始终保存地址，而 MDR 保存实际数值。PC 也保存地址，但它是专门保存下一条指令的地址。

When describing the fetch‑decode‑execute cycle, avoid vague phrases like ‘the instruction is fetched’. Always state which register supplies the address, how the data moves (via the address bus and data bus) and what happens to the PC. Marks are awarded for precise register naming.

在描述取指–解码–执行周期时，避免使用诸如“指令被取出”这样含糊的表述。一定要说明哪个寄存器提供地址，数据如何移动（通过地址总线和数据总线），以及 PC 发生了什么变化。准确地命名寄存器才能得分。

For performance questions, link each factor to the cycle. Clock speed directly affects how quickly cycles repeat. Cores allow true simultaneous execution of separate threads. Cache reduces the average time the CPU waits for data, thereby increasing overall throughput. Give concrete numerical examples where helpful.

对于性能相关问题，将每个因素与周期联系起来。时钟速度直接影响周期重复的速度。多个核心允许多个线程真正同时执行。高速缓存缩短了 CPU 等待数据的平均时间，从而提高了整体吞吐量。必要时可给出具体的数值示例。

Finally, make sure you can draw and label a simple Von Neumann diagram showing the CPU (with internal registers), the system bus and memory. Even a quick sketch in a written exam can earn several marks.

最后，请确保你能够画出并标注一个简单的冯·诺依曼架构图，展示 CPU（及其内部寄存器）、系统总线和存储器。在笔试中哪怕是快速的草图也能为你赢得若干分数。

10. CCEA-Style Quick Recap and Revision Checklist | CCEA 风格快速回顾与复习清单

Use this checklist to verify your readiness:

请使用以下检查清单验证你的备考情况：

Can you name all the Von Neumann components? 能否说出所有冯·诺依曼架构的组件？
Do you know the roles of PC, MAR, MDR, CIR, ACC? 是否了解 PC、MAR、MDR、CIR、ACC 的作用？
Can you explain the three types of bus and their direction? 能否解释三种总线类型及其方向？
Can you step through the fetch‑decode‑execute cycle with register transfers? 能否借助寄存器传输逐步讲解取指–解码–执行周期？
What is the Von Neumann bottleneck and how does cache help? 什么是冯·诺依曼瓶颈？高速缓存如何缓解它？
How do clock speed, cores and cache each affect performance? 时钟速度、核心数和缓存分别如何影响性能？
What is the difference between volatile and non‑volatile storage? 易失性存储与非易失性存储有何区别？
Can you describe an embedded system and give two real‑world examples? 能否描述嵌入式系统并给出两个现实世界的例子？

Master these bullet points and you will be well prepared for any architecture question on the IGCSE CCEA Computer Science paper.

掌握以上要点，你就能从容应对 IGCSE CCEA 计算机科学试卷中任何一道体系结构考题。

Published by TutorHao | Computer Science Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Computer Architecture Essentials for IGCSE CCEA Computer Science | 计算机体系结构考点精讲