A-Level CCEA Computer Science: Memory Essentials | A-Level CCEA 计算机：存储器考点精讲

📚 A-Level CCEA Computer Science: Memory Essentials | A-Level CCEA 计算机：存储器考点精讲

Memory is a cornerstone of computer architecture, directly influencing how data is stored, accessed, and processed. This article explores the essential memory concepts required for the CCEA A-Level Computer Science specification, from primary memory hierarchies to virtual memory and secondary storage technologies.

存储器是计算机体系结构的基石，直接影响数据的存储、访问和处理方式。本文深入讲解 CCEA A-Level 计算机科学考试中必备的存储器核心概念，涵盖主存层次结构、虚拟内存以及辅存技术。

1. The Memory Hierarchy | 存储器层次结构

The memory hierarchy organises storage by speed, cost, and capacity. Registers sit at the top, offering the fastest access but minimal capacity, followed by cache, main memory (RAM), and secondary storage. This structure balances performance and cost, leveraging the principle of locality to keep frequently used data closer to the CPU.

存储器层次结构根据速度、成本和容量来组织存储。寄存器位于顶端，速度最快但容量极小，随后是高速缓存、主存（RAM）和辅存。这种结构平衡了性能与成本，利用局部性原理将频繁使用的数据放在离 CPU 更近的地方。

Registers: built into the CPU; fastest but smallest capacity (a few bytes).
Cache: small, fast SRAM; stores copies of frequently accessed main memory data.
Main memory (RAM): holds currently executing programs and data; larger but slower than cache.
Secondary storage: non-volatile, high capacity, slowest; e.g. HDD, SSD.

寄存器：集成在 CPU 内部；速度最快，容量最小（几个字节）。
高速缓存（Cache）：小而快的 SRAM；保存常用主存数据的副本。
主存（RAM）：存放正在执行的程序和数据；容量较大但比缓存慢。
辅存：非易失，容量大，速度最慢；如 HDD、SSD。

2. Primary Memory: RAM and ROM | 主存储器：RAM 与 ROM

Primary memory is directly accessible by the CPU. The two fundamental types are RAM (Random Access Memory) and ROM (Read-Only Memory). RAM is volatile and used for temporary storage of instructions and data during execution; ROM is non-volatile and typically stores firmware or boot routines.

主存储器可被 CPU 直接访问。两种基本类型是 RAM（随机存取存储器）和 ROM（只读存储器）。RAM 是易失性的，用于在执行期间临时存储指令和数据；ROM 是非易失性的，通常存储固件或引导程序。

RAM can be static (SRAM) or dynamic (DRAM). SRAM uses flip-flops, is faster and more expensive, and is used for cache. DRAM uses capacitors, needs refreshing, is slower but cheaper, and is used for main memory.

RAM 分为静态（SRAM）和动态（DRAM）。SRAM 使用触发器，速度更快、价格更高，用于高速缓存。DRAM 使用电容，需要刷新，速度较慢但更便宜，用于主存。

ROM variants include PROM, EPROM, and EEPROM, which differ in how they can be programmed and erased. In modern systems, flash memory (a type of EEPROM) is widely used for firmware.

ROM 的变体包括 PROM、EPROM 和 EEPROM，区别在于编程和擦除方式的不同。现代系统中，闪存（一种 EEPROM）广泛用于固件存储。

3. DRAM and the Refreshing Mechanism | DRAM 与刷新机制

Dynamic RAM stores each bit as a charge on a tiny capacitor. Because the charge leaks away, DRAM cells must be periodically refreshed—read and rewritten—to retain data. This refreshing process consumes power and time, contributing to DRAM’s slower speed compared to SRAM.

动态 RAM 将每个比特存储为微小电容上的电荷。由于电荷会泄漏，DRAM 单元必须定期刷新（读取并重新写入）以保持数据。这一刷新过程消耗功耗和时间，因此 DRAM 比 SRAM 慢。

The refresh cycle is managed by the memory controller and typically happens every few milliseconds. During refresh, part of the memory may be temporarily unavailable, which can slightly impact overall system performance.

刷新周期由内存控制器管理，通常每隔几毫秒进行一次。刷新期间，部分内存可能暂时不可用，这会轻微影响整体系统性能。

4. SRAM Technology and Applications | SRAM 技术与应用

Static RAM uses a latch circuit (typically six transistors per bit) to hold data, so it does not need refreshing. This makes SRAM faster and more power-efficient in idle mode, but its complexity results in lower density and higher cost per bit.

静态 RAM 使用锁存电路（每比特通常六个晶体管）保存数据，因此无需刷新。这使得 SRAM 速度更快、空闲模式下功耗更低，但其复杂性导致集成度较低、每比特成本更高。

SRAM is primarily used for CPU caches (L1, L2, L3) where speed is critical. It can also appear in battery-backed storage for settings, but its high cost prevents it from being used as main memory in consumer devices.

SRAM 主要用于 CPU 高速缓存（L1、L2、L3），这些地方对速度要求极高。它也可能出现在带电池备份的设置存储中，但高成本使其无法在消费设备中用作主存。

5. Cache Memory: Levels and Operation | 高速缓存：层级与工作方式

Cache is a small amount of high-speed SRAM located close to the CPU. It stores frequently accessed instructions and data to reduce the average time to access main memory. Modern processors typically have multiple levels: L1 (fastest, smallest), L2, and L3 (larger, shared).

缓存是位于 CPU 附近的小容量高速 SRAM。它存储频繁访问的指令和数据，以减少访问主存的平均时间。现代处理器通常有多级缓存：L1（最快、最小）、L2 和 L3（更大、共享）。

When the CPU requests data, the cache controller checks if it is present (a hit) or not (a miss). A miss triggers a fetch from main memory, and the data is placed into the cache according to a replacement policy (e.g., LRU). The hit rate significantly affects performance.

当 CPU 请求数据时，缓存控制器检查数据是否存在（命中）或不存在（缺失）。发生缺失时会从主存读取数据，并按替换策略（如 LRU）放入缓存。命中率对性能影响显著。

6. Virtual Memory and Paging | 虚拟内存与分页

Virtual memory allows the execution of programs that are larger than physical RAM by using a portion of secondary storage (e.g., HDD/SSD) as an extension. The OS divides memory into fixed-size blocks called pages, mapping virtual addresses to physical frames.

虚拟内存通过使用部分辅存（如 HDD/SSD）作为扩展，使得可以运行比物理 RAM 更大的程序。操作系统将内存划分为固定大小的块，称为页，并将虚拟地址映射到物理帧。

The page table stores the mappings. When a page is accessed but not in RAM (a page fault), the OS swaps it in from disk, possibly writing another page out to disk if needed. This swapping can cause disk thrashing if too many page faults occur, severely degrading performance.

页表存储映射关系。当访问的页不在 RAM 中（发生缺页），操作系统从磁盘调入该页，如有必要还会将另一页换出到磁盘。如果缺页过多，这种交换会导致磁盘抖动，严重降低性能。

7. Secondary Storage: Magnetic, Optical, Solid State | 辅存：磁、光、固态

Secondary storage provides non-volatile, long-term storage. Common types include magnetic hard disk drives (HDD), optical discs (CD, DVD, Blu-ray), and solid-state drives (SSD). They differ in access mechanisms, speed, durability, and cost per gigabyte.

辅存提供非易失的长期存储。常见类型包括磁性硬盘驱动器（HDD）、光盘（CD、DVD、蓝光）和固态硬盘（SSD）。它们在访问机制、速度、耐用性和每 GB 成本方面各不同。

HDDs use spinning platters and read/write heads; data access time includes seek time and rotational latency. SSDs employ NAND flash memory, offering much faster random access and lower power consumption, but have a limited number of write cycles.

HDD 使用旋转盘片和读写头；数据访问时间包括寻道时间和旋转延迟。SSD 采用 NAND 闪存，提供快得多的随机访问和更低功耗，但写入次数有限。

Optical storage uses laser light to read and write data on reflective surfaces. It is often used for media distribution and archiving, but its capacity and speed are generally lower than magnetic or solid-state alternatives.

光存储使用激光在反射面上读写数据。通常用于媒体分发和存档，但其容量和速度通常低于磁性或固态替代方案。

8. Memory Addressing and Address Bus | 存储器寻址与地址总线

The CPU communicates with memory via the address bus, data bus, and control bus. The width of the address bus determines the maximum number of memory locations that can be directly addressed. For example, an n-bit address bus can address 2ⁿ distinct memory locations.

CPU 通过地址总线、数据总线和控制总线与存储器通信。地址总线的宽度决定了可直接寻址的存储器位置最大数量。例如，n 位地址总线可寻址 2ⁿ 个不同的存储单元。

If a system has a 32-bit address bus, it can theoretically address 2³² bytes = 4 GiB of memory. Techniques like bank switching or PAE can extend this limit, but the fundamental relationship is key for understanding system limitations.

如果系统有 32 位地址总线，理论上可寻址 2³² 字节 = 4 GiB 内存。存储体切换或 PAE 等技术可以突破这一限制，但理解这一基本关系对掌握系统局限性至关重要。

Memory addresses refer to individual bytes, but data is often transferred in words (e.g., 4 bytes). The data bus width determines how many bits can be transferred simultaneously, influencing overall memory bandwidth.

存储器地址指向单个字节，但数据通常按字传输（例如 4 字节）。数据总线宽度决定了一次可传输多少位，影响整体内存带宽。

9. The Stored Program Concept and Von Neumann Architecture | 存储程序概念与冯·诺依曼架构

The stored program concept, fundamental to the Von Neumann architecture, places both program instructions and data in the same memory space. This unified memory allows the CPU to fetch and execute instructions sequentially, but it also creates the Von Neumann bottleneck—where the single bus between CPU and memory limits throughput.

存储程序概念是冯·诺依曼架构的基础，它将程序指令和数据存放在同一存储空间中。这种统一内存允许 CPU 顺序取指和执行指令，但也产生了“冯·诺依曼瓶颈”——CPU 与存储器之间的单一总线限制了吞吐量。

In contrast, Harvard architecture uses separate memory and buses for instructions and data, enabling simultaneous access. Some modern processors use a modified Harvard architecture, where caches are split but main memory is unified.

相对地，哈佛架构为指令和数据使用独立的存储器和总线，可以同时访问。一些现代处理器采用改进的哈佛架构，缓存分离但主存统一。

10. Memory Performance Metrics and Calculations | 存储器性能指标与计算

A-Level questions often involve calculating memory performance. Key metrics include access time (latency), cycle time, data transfer rate (bandwidth), and capacity. For example, given a memory with an 8-byte data bus operating at 200 MHz, the theoretical bandwidth is 8 bytes × 200 × 10⁶ Hz = 1600 MB/s.

A-Level 考试题常常涉及存储器性能计算。关键指标包括访问时间（延迟）、周期时间、数据传输率（带宽）和容量。例如，给定一个数据总线为 8 字节、工作于 200 MHz 的存储器，理论带宽为 8 字节 × 200 × 10⁶ Hz = 1600 MB/s。

Cache performance can be evaluated using average access time = hit time + miss rate × miss penalty. Understanding these formulas allows you to compare different memory designs and justify trade-offs in cost and performance.

缓存性能可用平均访问时间 = 命中时间 + 缺失率 × 缺失代价来评估。理解这些公式有助于比较不同的存储器设计，并在成本与性能之间做出权衡判断。

Metric	Definition	Typical Unit
Access time	Time from read request to data availability	ns
Bandwidth	Max data transferred per second	MB/s or GB/s
Capacity	Total memory cells × bits per cell	Bytes (B), KiB, MiB, GiB

指标	定义	通常单位
访问时间	从读请求到数据就绪所需时间	纳秒 (ns)
带宽	每秒最大传输数据量	MB/s 或 GB/s
容量	总存储单元数 × 每单元位数	字节 (B)，KiB，MiB，GiB

11. Modern Storage Technologies and Trends | 现代存储技术与趋势

NAND flash memory, used in SSDs, has evolved with 3D stacking (V-NAND) increasing density and lowering cost. New non-volatile memory technologies like Intel’s Optane (3D XPoint) offer speeds closer to DRAM with persistence, potentially blurring the line between primary and secondary storage.

用于 SSD 的 NAND 闪存已发展为 3D 堆叠（V-NAND），提高了集成度并降低了成本。新型非易失存储器技术如 Intel 的 Optane（3D XPoint）提供了接近 DRAM 的速度，同时具有持久性，可能模糊主存与辅存的界限。

Cloud and network-attached storage (NAS) are increasingly relevant, though they fall outside the direct A-Level scope. Understanding RAID levels (0, 1, 5, 10) can also appear in questions about reliability and performance in storage systems.

云存储和网络附加存储（NAS）日益重要，尽管不在 A-Level 直接测试范围内。了解 RAID 级别（0、1、5、10）也可能出现在有关存储系统可靠性和性能的考题中。

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

更多咨询请联系16621398022（同微信）

A-Level CCEA Computer Science: Memory Essentials | A-Level CCEA 计算机：存储器 考点精讲