📚 IB & Edexcel Computer Science: Mind Map Quick Revision | IB 与 Edexcel 计算机科学：思维导图速记
Mind maps are powerful visual tools that help you organise and memorise key computing concepts for IB and Edexcel Computer Science exams. This revision guide uses a mind map approach to cover essential topics, grouping related ideas into branches for quick recall. By linking concepts logically, you can reinforce your understanding and improve exam performance.
思维导图是强大的视觉工具，可帮助你组织和记忆 IB 与 Edexcel 计算机科学考试的关键概念。本复习指南采用思维导图法覆盖核心主题，将相关知识点按分支分组以便快速回忆。通过逻辑关联概念，你能加深理解并提高考试成绩。
Each section below is written as one branch of a master mind map: a central idea, a few sub-branches, and the exact vocabulary examiners reward. Read the English, then the Chinese, then close the page and redraw the branch from memory. That single habit - draw, check, correct - is worth more than re-reading any textbook.
下面每一节都相当于一张总思维导图的一个分支：一个中心概念、几个子分支，以及考官真正给分的专业术语。先读英文，再读中文，然后合上页面凭记忆重画这个分支。这个"画—核对—修正"的习惯，比反复阅读任何教材都更有效。
1. Why Mind Maps Work for Computer Science | 为什么思维导图适合计算机科学
A mind map works because it matches how memory is stored: in hierarchies of chunks, not in long paragraphs. Computer Science is unusually well suited to this because almost every topic is already a hierarchy - a system contains components, a component contains registers, a register holds a value. A mind map simply makes that existing structure visible.
思维导图之所以有效，是因为它符合记忆的存储方式：以层级化的"组块"存在，而非长段落。计算机科学尤其适合这种方法，因为几乎每个主题本身就是层级结构——系统包含部件，部件包含寄存器，寄存器保存数值。思维导图只是把这种既有结构可视化。
Two rules keep a computing mind map useful rather than decorative. First, keywords only: if a branch needs a full sentence, it is too long. Second, one colour per branch strand, and arrows for cross-links between strands, because exam questions rarely test a single topic in isolation.
有两条规则能让计算机思维导图保持实用而非装饰性。第一，只写关键词：如果一个分支需要完整句子，说明它太长了。第二，每个分支用不同颜色，并用箭头标出分支之间的横向联系，因为考题很少只孤立考查单一主题。
A typical master map for the whole course has six to eight top-level branches. The table below shows a workable structure and where each branch earns marks.
整门课程的总图通常有六到八个顶层分支。下表给出一个可行的结构，以及每个分支的得分点。
| Branch | 分支 | Core content | 核心内容 | Typical question style | 典型题型 |
| Hardware | 硬件 | CPU, registers, buses, memory, storage | CPU、寄存器、总线、内存、存储 | Describe the fetch-decode-execute cycle | 描述取指-译码-执行周期 |
| Data | 数据 | Binary, hex, two's complement, bitmaps, sound | 二进制、十六进制、补码、位图、声音 | Calculate file size | 计算文件大小 |
| Networks | 网络 | Topologies, protocols, layers, addressing | 拓扑、协议、分层、寻址 | Explain packet switching | 解释分组交换 |
| Programming | 编程 | Constructs, data structures, OOP, translators | 结构、数据结构、面向对象、翻译程序 | Write pseudocode / trace a table | 写伪代码 / 填跟踪表 |
| Algorithms | 算法 | Searching, sorting, recursion, complexity | 查找、排序、递归、复杂度 | Compare efficiency of two algorithms | 比较两种算法的效率 |
| Issues | 议题 | Ethics, legislation, digital divide | 伦理、法律、数字鸿沟 | Discuss one benefit and one drawback | 讨论一个优点和一个缺点 |
2. Computer Systems and Architecture | 计算机系统与架构
The central processing unit (CPU) is the core of any computer system. It consists of the Control Unit (CU), which directs operations and manages the fetch-decode-execute cycle; the Arithmetic Logic Unit (ALU), which performs calculations and logical operations; and various registers. The fetch-decode-execute cycle begins with the Program Counter (PC) holding the address of the next instruction. The address is moved to the Memory Address Register (MAR), and the instruction is fetched into the Memory Data Register (MDR). It is then copied to the Current Instruction Register (CIR) for decoding. The CU decodes the instruction and coordinates the execution, often using the Accumulator (ACC) to store intermediate results. Clock speed, measured in GHz, determines how many cycles are executed per second, directly affecting performance.
中央处理器 (CPU) 是任何计算机系统的核心。它由控制单元 (CU)、算术逻辑单元 (ALU) 和各种寄存器组成。控制单元指导操作并管理取指-译码-执行周期；ALU 执行计算和逻辑运算。取指-译码-执行周期从程序计数器 (PC) 开始，它保存下一条指令的地址。该地址被传送到内存地址寄存器 (MAR)，指令被读取到内存数据寄存器 (MDR)。然后它被复制到当前指令寄存器 (CIR) 进行译码。CU 译码指令并协调执行，通常使用累加器 (ACC) 存储中间结果。以 GHz 为单位的时钟速度决定每秒执行多少个周期，直接影响性能。
Fetch → Decode → Execute → (repeat) | 1 GHz = 10⁹ cycles per second
→ decoded by CU → executed by ALU → result in ACC. If you can narrate that chain, you can answer almost any architecture question.
寄存器是计算机中最快的存储，每一个只负责一件事。请把它们当作"数据流动的链条"来记，而不是孤立清单：PC → MAR → MDR → CIR → 由 CU 译码 → 由 ALU 执行 → 结果存入 ACC。只要能复述这条链，几乎任何体系结构题都能作答。
Three buses connect the components: the address bus carries the location of data and is one-directional, the data bus carries the data itself and is bidirectional, and the control bus carries timing and command signals. Performance is not only clock speed: cache size, number of cores, word length and bus width all matter, and this trade-off is a favourite exam discussion point.
三种总线连接各部件：地址总线传输数据的位置，是单向的；数据总线传输数据本身，是双向的；控制总线传输时序与命令信号。性能不只取决于时钟速度：缓存大小、核心数量、字长和总线宽度都很关键，而这种权衡是考试中常见的讨论点。
The Von Neumann architecture shares one memory for instructions and data, which creates the Von Neumann bottleneck; the Harvard architecture uses separate instruction and data memories and is common in embedded systems. Pipelining improves throughput by overlapping the stages of several instructions at once.
冯·诺依曼架构的指令与数据共用同一存储器，因此形成"冯·诺依曼瓶颈"；哈佛架构使用独立的指令存储器与数据存储器，常见于嵌入式系统。流水线技术通过让多条指令的不同阶段重叠执行来提高吞吐量。
3. Memory, Storage and Data Representation | 内存、存储与数据表示
Primary memory is directly accessible by the CPU. RAM is volatile and holds the currently running programs and data; ROM is non-volatile and typically holds the bootstrap loader. Cache is small, fast memory close to the CPU (L1, L2, L3) that stores frequently used instructions, and virtual memory extends RAM by using part of secondary storage, managed through paging and page swapping.
主存可被 CPU 直接访问。RAM 是易失的，保存当前运行的程序和数据；ROM 是非易失的，通常保存引导加载程序。缓存是靠近 CPU 的小容量高速存储器（L1、L2、L3），存放频繁使用的指令；虚拟内存通过使用部分二级存储来扩展 RAM，由分页与页面交换机制管理。
Secondary storage is compared on five criteria: speed, capacity, durability, portability and cost per gigabyte. Magnetic disks are cheap per GB with large capacity; optical media are portable and good for archiving but slow; solid state drives are fast, silent and shock-resistant but more expensive per GB and have a finite number of write cycles.
二级存储通常从五个维度比较：速度、容量、耐用性、便携性和每 GB 成本。磁介质每 GB 便宜、容量大；光介质便携、适合归档但速度慢；固态硬盘速度快、无噪音、抗震，但每 GB 更贵，且写入次数有限。
1 KiB = 2¹⁰ bytes = 1024 bytes | 1 MiB = 2²⁰ bytes | 1 GiB = 2³⁰ bytes
For data representation, remember the conversion triangle: binary ↔ denary ↔ hexadecimal. Hexadecimal is used because it is compact and each hex digit maps cleanly to four bits, which makes memory addresses and colour codes readable. Two's complement allows negative numbers in binary: invert the bits and add one.
在数据表示方面，记住"转换三角"：二进制 ↔ 十进制 ↔ 十六进制。使用十六进制是因为它紧凑，且每个十六进制位恰好对应四位二进制，便于阅读内存地址和颜色代码。补码用二进制表示负数：按位取反再加一。
Two's complement range for n bits: −2ⁿ⁻¹ to 2ⁿ⁻¹ − 1
Character sets: ASCII uses 7 bits and covers 128 characters, which is enough for English but not for global scripts; Unicode uses more bits and supports thousands of characters and emoji, at the cost of larger file sizes. Bitmap images are calculated from resolution and colour depth, while vector images store geometric objects and scale without loss of quality.
字符集方面：ASCII 使用 7 位，可表示 128 个字符，对英语足够，但无法覆盖全球文字；Unicode 使用更多位，支持成千上万个字符和表情符号，代价是文件更大。位图图像由分辨率和色深计算大小，而矢量图存储几何对象，缩放不损失画质。
Bitmap size (bits) = width × height × colour depth | Sound size (bits) = sample rate × bit depth × duration × channels
Sound is digitised by sampling: a higher sample rate and greater bit depth give better fidelity and a larger file. Compression is either lossless (original data fully recoverable, e.g. PNG, FLAC) or lossy (data permanently discarded for a much smaller file, e.g. JPEG, MP3).
声音通过采样被数字化：采样率越高、位深越大，音质越好，文件也越大。压缩分为无损（可完全还原原始数据，如 PNG、FLAC）和有损（永久丢弃数据以换取更小文件，如 JPEG、MP3）。
4. Operating Systems and System Software | 操作系统与系统软件
The operating system is the layer between hardware and user. Its core jobs are memory management, process scheduling, handling interrupts, file management, device driver communication, providing a user interface, and enforcing security and user accounts. Be ready to give a concrete example of each, because "explain the role of the OS" marks are usually awarded per function.
操作系统是硬件与用户之间的中间层。其核心职责包括内存管理、进程调度、中断处理、文件管理、设备驱动通信、提供用户界面，以及实施安全与用户账户管理。要能为每项职责举出具体例子，因为"解释操作系统作用"这类题通常按功能点给分。
Scheduling decides which process gets the CPU next. First come first served is simple but suffers from the convoy effect; shortest job first minimises average waiting time but can starve long jobs; round robin gives each process a fixed time slice and suits time-sharing systems; priority scheduling favours high-priority tasks and may need ageing to avoid starvation.
调度决定下一个获得 CPU 的进程。先来先服务简单，但会出现"护航效应"；最短作业优先可让平均等待时间最小，但可能使长作业饿死；时间片轮转给每个进程固定时间片，适合分时系统；优先级调度偏向高优先级任务，可能需要"老化"机制防止饿死。
An interrupt is a signal that temporarily halts the current process so a higher-priority task can be served. The processor saves the current state to the stack, looks up the interrupt service routine (ISR) in the interrupt vector table, runs it, then restores the saved state. Polling, by contrast, means the CPU repeatedly checks a device, which wastes cycles.
中断是一种信号，会暂停当前进程以便先处理更高优先级的任务。处理器先把当前状态保存到栈中，再于中断向量表中查找中断服务程序 (ISR)，执行完毕后再恢复保存的状态。与此相对，轮询指 CPU 反复检查某设备，会浪费时钟周期。
Utility software keeps the system healthy: defragmentation, backup, compression, anti-malware, encryption and disk formatting. Note the exam distinction - the OS manages the system, utilities maintain it, and application software serves the user.
实用程序用于维护系统健康：磁盘碎片整理、备份、压缩、反恶意软件、加密和磁盘格式化。注意考试中的区分：操作系统管理系统，实用程序维护系统，应用软件服务用户。
5. Networks and the Internet | 网络与互联网
A network is two or more devices connected to share data and resources. A LAN covers a small area and is usually owned by one organisation; a WAN spans large distances and typically uses infrastructure owned by third parties. A PAN is a personal network around one person, such as Bluetooth headphones connected to a phone.
网络是两台或更多设备相互连接以共享数据和资源。局域网 (LAN) 覆盖小范围，通常由单一机构拥有；广域网 (WAN) 跨越远距离，一般使用第三方拥有的基础设施。个人区域网 (PAN) 是围绕个人的网络，例如蓝牙耳机连接手机。
Client-server networks centralise control, backup and security but need a skilled administrator and cost more; peer-to-peer networks are cheap and easy to set up but offer weaker security and no central backup. Topology matters too: a star network is easy to fault-find and expand but fails if the central switch fails; a mesh network is highly resilient with many redundant paths but is expensive to cable.
客户端-服务器网络把控制、备份和安全集中化，但需要专业管理员且成本更高；对等网络 (P2P) 便宜且易搭建，但安全性弱、没有集中备份。拓扑结构同样重要：星型网络易于查错和扩展，但中心交换机故障会导致全网瘫痪；网状网络因有多条冗余路径而非常可靠，但布线成本高昂。
Packet switching breaks data into packets, each containing the destination address, source address, sequence number and payload. Routers forward packets independently along the cheapest available route, and the receiving device reorders them; this is efficient and resilient, but packets can arrive out of order or be lost, so TCP is used to request retransmission.
分组交换把数据拆成多个分组，每个分组包含目标地址、源地址、序号和有效载荷。路由器沿当前最经济的路径独立转发分组，接收端再重新排序；这种方式高效且可靠，但分组可能乱序或丢失，因此用 TCP 请求重传。
Memorise protocols by function, not alphabetically: HTTP/HTTPS for web pages and secure web pages, FTP for file transfer, SMTP for sending mail, IMAP and POP3 for retrieving mail, TCP for reliable ordered delivery, UDP for fast low-overhead streaming, IP for addressing and routing, DNS for translating domain names into IP addresses, and DHCP for automatic IP assignment.
协议要按功能记忆，而不是按字母顺序：HTTP/HTTPS 用于网页与安全网页，FTP 用于文件传输，SMTP 用于发送邮件，IMAP 与 POP3 用于接收邮件，TCP 提供可靠有序的传输，UDP 用于快速低开销的流媒体，IP 负责寻址与路由，DNS 把域名翻译成 IP 地址，DHCP 自动分配 IP 地址。
Every networked device has a MAC address (a permanent hardware identifier) and an IP address (a logical, changeable address). IPv4 uses 32 bits and is running out
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图