OCR A Level Computer Science Paper 1 — Last-Minute Revision Guide (H446/01)
Everything you need to know for OCR A Level Computer Science Component 1 exam. CPU architecture, memory, networks, databases, Boolean logic, data representation and ethics — with exam tips and mark-scheme language.
Gareth Edgell
Head of CS · Senior Examiner · 15+ years tutoring
OCR A Level Computer Science Component 1 (H446/01) covers the theory side of Computer Science — from CPU architecture to databases, Boolean logic to ethical issues. It is a 2.5-hour written paper worth 40% of your A Level.
This guide covers every major topic with the exact wording and detail the mark scheme rewards.
1. The Processor (CPU Architecture)
Von Neumann architecture
The Fetch-Decode-Execute (FDE) cycle is the most examined topic in this section.
Fetch:
- The address in the Program Counter (PC) is copied to the Memory Address Register (MAR)
- The instruction at that address is copied to the Memory Data Register (MDR)
- The PC is incremented (ready for the next instruction)
- The instruction is copied from the MDR to the Current Instruction Register (CIR)
Decode:
- The Control Unit (CU) decodes the instruction in the CIR
- It splits the instruction into opcode (what to do) and operand (data/address)
Execute:
- The Arithmetic Logic Unit (ALU) performs calculations or logical operations
- Results are stored back to the MDR/registers
Exam tip: “Describe the FDE cycle” is worth 4–6 marks. Name each register by its full name, state what gets copied from and to, and remember PC increments during Fetch.
Registers you must know
| Register | Full Name | Purpose |
|---|---|---|
| PC | Program Counter | Holds address of NEXT instruction |
| MAR | Memory Address Register | Holds address to read from/write to memory |
| MDR | Memory Data Register | Holds data being transferred to/from memory |
| CIR | Current Instruction Register | Holds the instruction being executed |
| ACC | Accumulator | Holds results of ALU operations |
| IX | Index Register | Used in indexed addressing |
Factors affecting CPU performance
- Clock speed — more cycles per second = faster. Measured in GHz. Doubling clock speed roughly doubles performance.
- Number of cores — each core can process instructions independently. 4 cores ≠ 4× faster (Amdahl’s Law — some tasks can’t be parallelised).
- Cache size — L1 (fastest, smallest), L2, L3. Data in cache is accessed far faster than RAM. Larger cache = fewer cache misses.
- Pipelining — fetching the next instruction while executing the current one. Increases throughput but hazards (data/control) can stall the pipeline.
CISC vs RISC
| CISC | RISC | |
|---|---|---|
| Instructions | Many complex instructions | Small set of simple instructions |
| Instruction size | Variable | Fixed (single clock cycle each) |
| Memory access | Many instructions access memory | Only load/store access memory |
| Examples | x86 (Intel/AMD PC) | ARM (mobile phones, Apple Silicon) |
2. Types of Memory
RAM vs ROM
- RAM (Random Access Memory) — volatile, read/write, stores running programs and data
- ROM (Read Only Memory) — non-volatile, read only, stores BIOS/firmware
Virtual memory
When RAM is full, the OS uses part of the hard drive as additional memory. This is called virtual memory. Pages of data are swapped between RAM and disk. It prevents programs crashing when RAM runs out, but is much slower than RAM (disk access ≈ 100,000× slower).
Thrashing: when the CPU spends more time swapping pages than executing instructions. Caused by too many processes for available RAM.
Cache memory
Sits between the CPU and RAM. Much faster than RAM (nanoseconds vs tens of nanoseconds). The CPU checks cache before RAM:
- Cache hit — data found in cache → very fast access
- Cache miss — data not in cache → must fetch from RAM (slower), then cache it
Spatial locality: recently accessed data tends to be near other soon-to-be-accessed data — so the cache loads a block of memory around the accessed address.
3. Secondary Storage
Magnetic (HDD)
- Spinning platters, read/write heads
- Slower than SSD but cheaper per GB
- Good for large file storage; bad for OS drives
Solid State (SSD)
- Flash memory (NAND), no moving parts
- Much faster than HDD, more durable, silent
- More expensive per GB; finite write cycles
Optical (CD/DVD/Blu-ray)
- Laser reads/writes pits and lands on a disc
- CD: ~700MB, DVD: ~4.7GB, Blu-ray: ~25GB
- Used for distribution, archiving, backups
Exam tip: “Give two reasons why an SSD is more suitable than an HDD for a laptop” — accept: faster boot/load times, no moving parts (more robust to knocks), lighter, silent, lower power consumption.
4. Operating Systems
Functions of an OS
- Process management — scheduling CPU time between processes (FCFS, Round Robin, SJF, Priority)
- Memory management — allocating RAM to processes, virtual memory, paging
- File management — organising files in directories, permissions, formats
- Device management — drivers that abstract hardware from applications
- Security management — user authentication, access control
Scheduling algorithms
| Algorithm | Description | Advantage | Disadvantage |
|---|---|---|---|
| FCFS | First Come First Served | Simple | Long jobs block short ones |
| Round Robin | Each process gets a time slice | Fair, responsive | Overhead from context switching |
| SJF | Shortest Job First | Minimises average wait | Can starve long jobs |
| Priority | Highest priority first | Important jobs finish fast | Low-priority jobs may starve |
Types of OS
- Multi-tasking — appears to run multiple programs at once (rapid switching)
- Multi-user — multiple users share one system simultaneously
- Real-time — guaranteed response within a fixed time (medical, flight control)
- Distributed — workload spread across networked computers
- Embedded — dedicated, minimal OS in a device (washing machine, car ECU)
5. Boolean Logic and Logic Gates
Truth tables you must know
| A | B | AND | OR | NAND | NOR | XOR |
|---|---|---|---|---|---|---|
| 0 | 0 | 0 | 0 | 1 | 1 | 0 |
| 0 | 1 | 0 | 1 | 1 | 0 | 1 |
| 1 | 0 | 0 | 1 | 1 | 0 | 1 |
| 1 | 1 | 1 | 1 | 0 | 0 | 0 |
NOT: output = inverse of input (1-bit only)
De Morgan’s Laws
¬(A ∧ B) ≡ ¬A ∨ ¬B — NOT AND = OR of NOTs
¬(A ∨ B) ≡ ¬A ∧ ¬B — NOT OR = AND of NOTs
Simplification — key Boolean identities
- A AND 1 = A
- A AND 0 = 0
- A OR 1 = 1
- A OR 0 = A
- A AND A = A
- A OR A = A
- A AND NOT A = 0
- A OR NOT A = 1
- NOT (NOT A) = A
Karnaugh maps
Group 1s in groups of 1, 2, 4 or 8. Each group gives one term. Wrap-around edges count. Larger groups = simpler expression.
Exam tip: On a 6-mark Boolean simplification question, show EVERY step citing the law used. Don’t skip steps — each step can earn a mark.
6. Data Representation
Number bases
Binary to denary: multiply each bit by its place value (128, 64, 32, 16, 8, 4, 2, 1) and add up.
Denary to binary: repeated division by 2, read remainders bottom to top.
Binary to hex: group bits in 4s from the right, convert each group (0–9, A–F).
Two’s complement
Used to represent negative integers.
To negate: flip all bits, then add 1.
Example: −35 in 8-bit two’s complement
- 35 = 00100011
- Flip: 11011100
- Add 1: 11011101
To read a negative two’s complement: the MSB has negative place value (−128 for 8-bit).
Floating point
IEEE 754 single precision: 1 sign bit, 8 exponent bits (biased by 127), 23 mantissa bits.
- Normalised: mantissa starts with 1.x for non-zero numbers (maximises precision)
- Overflow: result too large to represent
- Underflow: result too small (rounds to zero)
- Rounding error: many values can’t be represented exactly — accumulated errors can cause problems
Exam tip: “Give one disadvantage of using floating point” — rounding errors accumulate; limited precision; slower processing than integer arithmetic.
Sound representation
- Sample rate (Hz) — how many samples per second. Higher = more accurate. CD: 44,100 Hz.
- Bit depth — bits per sample. Higher = more dynamic range. CD: 16-bit.
- File size = sample rate × bit depth × duration (seconds) × channels
Image representation
- Pixel — smallest addressable element
- Resolution — total number of pixels (e.g., 1920 × 1080)
- Colour depth — bits per pixel. 24-bit = 16.7 million colours (8 bits each R, G, B)
- File size = width × height × colour depth (in bits), ÷ 8 for bytes
Compression
Lossless — original data fully recoverable. Run-length encoding, Huffman coding. Used for: text files, program files, PNG images.
Lossy — some data discarded permanently. JPEG images, MP3 audio. Smaller files but quality reduced. Humans don’t notice removed high-frequency detail.
7. Networks
Network protocols — the TCP/IP stack
| Layer | Protocol | Function |
|---|---|---|
| Application | HTTP, HTTPS, FTP, SMTP, DNS | Application-specific communication |
| Transport | TCP, UDP | Segmentation, error checking, flow control |
| Internet | IP | Routing, addressing (IP addresses) |
| Link | Ethernet, Wi-Fi | Physical transmission, MAC addresses |
TCP vs UDP:
- TCP: reliable, ordered, error-checked, connection-oriented. Used for web, email.
- UDP: fast, connectionless, no guarantee of delivery. Used for video streaming, DNS, gaming.
IP and MAC addresses
- IP address — logical address assigned by the network (can change). Used for routing across networks.
- MAC address — physical address burned into the NIC (permanent). Used for delivery within a network.
Encryption
- Symmetric — same key to encrypt and decrypt (AES). Fast but key exchange is a problem.
- Asymmetric — public key encrypts, private key decrypts (RSA). Solves key exchange problem. Slower.
- TLS/HTTPS — uses asymmetric to exchange a session key, then symmetric for the actual data.
Threats and countermeasures
| Threat | Description | Countermeasure |
|---|---|---|
| Phishing | Fake emails/sites to steal credentials | User education, spam filters |
| Malware | Software that harms a system | Anti-virus, firewalls, updates |
| DDoS | Flood server with requests | CDN, rate limiting, firewall rules |
| SQL injection | Malicious SQL in input fields | Parameterised queries, input validation |
| Man-in-the-middle | Intercept communication | Encryption (TLS/HTTPS), certificate pinning |
| Brute force | Try all password combinations | Account lockout, strong password policy |
8. Databases
Relational databases
A relational database stores data in tables (relations) with:
- Primary key — unique identifier for each row
- Foreign key — primary key of another table, creating a relationship
Normalisation removes data redundancy and update anomalies:
- 1NF — no repeating groups; each cell holds one value
- 2NF — 1NF + no partial dependencies on a composite key
- 3NF — 2NF + no transitive dependencies
SQL — key commands
SELECT name, age FROM Students WHERE age > 16 ORDER BY name;
SELECT Students.name, Courses.title
FROM Students
INNER JOIN Enrollments ON Students.id = Enrollments.student_id
INNER JOIN Courses ON Enrollments.course_id = Courses.id;
INSERT INTO Students (name, age) VALUES ('Alice', 17);
UPDATE Students SET age = 18 WHERE name = 'Alice';
DELETE FROM Students WHERE name = 'Bob';Exam tip: Most SQL questions involve a SELECT with WHERE and a JOIN. Practice writing the JOIN condition — it’s the most common error.
ACID properties
Transactions must be:
- Atomic — all or nothing (no partial transactions)
- Consistent — database remains valid before and after
- Isolated — concurrent transactions don’t interfere
- Durable — committed changes persist even after crash
9. Ethical, Legal and Cultural Issues
Legislation you must know
| Act | Key Provisions |
|---|---|
| Computer Misuse Act 1990 | Unauthorised access, unauthorised modification, DDoS attacks |
| Data Protection Act 2018 / GDPR | How personal data can be collected, stored, processed; data subject rights |
| Investigatory Powers Act 2016 | Government surveillance powers; bulk data collection |
| Copyright, Designs & Patents Act 1988 | Protects intellectual property; illegal to copy without licence |
Ethical frameworks
- Utilitarian — the greatest good for the greatest number
- Deontological — some actions are intrinsically right or wrong (Kantian ethics)
- Virtue ethics — what would a person of good character do?
Topics that come up every year
- AI and automated decision-making — bias in training data, accountability
- Environmental impact — energy use of data centres, e-waste
- Digital divide — unequal access to technology
- Privacy vs security — surveillance vs civil liberties
- Intellectual property — open source vs proprietary software
Last-minute exam strategy
- Read every question twice before writing anything
- Use technical vocabulary — mark schemes reward specific terms (e.g., “cache miss” not just “not in cache”)
- Match your answer length to the marks — a 4-mark question needs 4 distinct points
- For extended answers — use PEE structure: Point → Evidence → Explanation
- Don’t leave any question blank — attempt every question, even if unsure
- Check timing — 2.5 hours for the paper ≈ 1.5 minutes per mark
Good luck! If you want to practise any topic with past-paper style questions and instant AI marking, try the Question Bank on CompSci Tutoring.