To learn Cambridge IGCSE Computer Science (0478) quickly and correctly, stop treating computers as mysterious boxes and programming as memorising commands. The subject asks how data is represented and processed, how digital systems work, and how you can design an accurate solution to a problem. This eduKateSG Quick Notes guide follows the 2026–2028 Cambridge syllabus. You will learn the ten topics, the two examination papers, and a small daily method for explaining a concept, tracing an algorithm and checking that your result is correct.
- Understand before memorising: ask what goes in, what happens and what comes out.
- Connect binary to real data: practise converting numbers and explaining how text, images and sound are represented.
- Practise systems questions: learn hardware, software, networks, security and emerging technologies through clear scenarios.
- Write algorithms in steps: identify inputs, processing, outputs and any storage.
- Trace and test: use example values to discover logic errors rather than guessing.
- Follow your examination year: 0478 in 2026–2028 uses pseudocode, a 15-mark scenario question and two non-calculator papers.
First, imagine an ordinary task: a teacher wants the computer to calculate the average of three quiz scores. You already know the human method. Add the three numbers, then divide by three. Now turn it into a computer solution: accept three inputs, calculate the total, calculate the average and output the result. That is input → processing → output. The device can execute the steps, but you must explain them precisely enough that every valid input gives the intended result.
Second, test the steps with simple values. If the scores are 6, 9 and 12, the total is 27 and the average is 9. A computer program that prints 27 has calculated the total but has not answered the user’s question. A program that divides by the number of students when that number is zero needs special handling. Computer Science rewards both a working result and an explanation of why the logic works. Learn to spot the error by tracing variables one step at a time.
Third, practise transferring the same logic to a new situation. Replace quiz scores with three temperatures, three journey times or three item prices. Which parts of the algorithm stay unchanged? Which parts might need different units, precision or input checks? When you can adapt the method instead of copying it blindly, you have learned computational thinking. This is the purpose of the subject: breaking a problem into manageable decisions and checking each decision.
The 2026–2028 syllabus: ten topics in two connected groups
Cambridge separates the syllabus into Computer Systems (Topics 1–6) and Algorithms, Programming and Logic (Topics 7–10). Treat them as complementary. The first explains the capabilities and constraints of a computer; the second helps you create reliable instructions that use those capabilities.
- 1. Data representation: binary, denary and hexadecimal; representing text, sound and images; storage sizes and compression.
- 2. Data transmission: moving information between devices, errors and data security.
- 3. Hardware: processors, memory, input/output devices and storage.
- 4. Software: system software and applications, including how they support computer operations.
- 5. The internet and its uses: networked services, digital communication and security.
- 6. Automated and emerging technologies: how computing systems are used to solve practical problems.
- 7. Algorithm design and problem-solving: decomposition, flowcharts, pseudocode, validation and testing.
- 8. Programming: variables, data types, selection, iteration, arrays, procedures and functions.
- 9. Databases: records, fields, keys and specified single-table SQL operations.
- 10. Boolean logic: truth tables, logic gates and interpreting logic circuits.
The two Cambridge examination papers
Paper 1 — Computer Systems: 1 hour 45 minutes, 75 marks and 50% of the overall qualification. It covers Topics 1–6 through compulsory short-answer and structured questions. Paper 2 — Algorithms, Programming and Logic: 1 hour 45 minutes, 75 marks and the other 50%. It covers Topics 7–10, including the scenario question. All candidates take both papers, and calculators are not permitted for either one.
Paper 2 has a final 15-mark unseen scenario question. Cambridge expects candidates to devote approximately 30 minutes to it. The 2026–2028 syllabus allows the solution for that specific question in Cambridge pseudocode or in Python, Visual Basic or Java. For other programming-solution questions requiring coding, the syllabus specifies pseudocode rather than programming-language code. This distinction is important: learning a language helps you test ideas, but you must also practise writing the notation the paper actually accepts.
Cambridge has separately published a 2029–2031 syllabus that changes the programming requirements and introduces a digital examination route under a different syllabus code. Do not apply those later changes to a 2026, 2027 or 2028 candidate. Ask your school which year and entry code you are following, then study its matching specification.
Worked example: binary is another way of writing a number
In base ten, the number 13 is one ten and three ones. In binary, each place is a power of two: 8, 4, 2, 1. Since 13 = 8 + 4 + 1, write a one under 8, 4 and 1 and zero under 2. The binary number is 1101, or 00001101 when shown in eight bits. Test the reverse conversion: 1×8 + 1×4 + 0×2 + 1×1 = 13.
Try a second example. The binary number 10110 has positions 16, 8, 4, 2 and 1. It represents 16 + 4 + 2 = 22. Don’t try to memorise whole conversion tables. Understand the place values and then practise converting in both directions.
Why eight-bit overflow occurs
An unsigned eight-bit register represents integers from 0 to 255. Suppose you try to add 250 and 12. The mathematical answer is 262, which is larger than 255. The value cannot fit in eight bits without losing information; this is an overflow condition. A simple unsigned eight-bit arithmetic result can wrap to 6, but the key examination idea is that the required number exceeds the register’s representable range. Always check how the question defines its representation.
Worked example: a short algorithm and trace table
Suppose a program starts with Total = 0 and repeatedly adds three inputs. The inputs are 4, 7 and 2. After the first input, Total is 4; after the second, 11; and after the third, 13. The algorithm then displays 13. A trace table should record how the variable changes at each step, not just the final value. Tracing is especially helpful when loops contain conditions or update several variables.
- Initial state: Total = 0.
- After input 4: Total = 4.
- After input 7: Total = 11.
- After input 2: Total = 13.
- Output: 13.
Now ask what happens if the inputs are −2, 5 and 0. The algorithm should output 3 if negative and zero inputs are permitted. If the task only allows positive scores, the program needs an appropriate validation rule. An algorithm must be judged against its stated requirements, not merely one example that happens to work.
Validation and verification are different
Validation checks whether entered data meet rules such as acceptable type, range or format. A score required to be between 0 and 100 can be checked with a range test. But 85 might pass validation even if the student actually scored 58. Verification tries to ensure data were entered or copied correctly, for example by double entry or careful visual checking. Passing validation is not proof of truth.
Boolean logic without confusion
An AND gate outputs 1 only when both inputs are 1. An OR gate outputs 1 when at least one input is 1. A NOT gate reverses its single input. For an AND gate with inputs A = 1 and B = 0, the output is 0. Start with these small cases, then expand into truth tables for the NAND, NOR and XOR gates listed in the syllabus.
When drawing a logic circuit, read the expression in the correct order and use the specified gate symbols. If there are three inputs, list every combination systematically. For example, three binary inputs have 2³ = 8 possible input rows. The easiest way to prevent a missing row is to write the binary combinations in order before calculating outputs.
How to think about databases
Imagine a table storing books. One field contains BookID, another contains Title and another Borrowed. Each row is a record. A unique BookID may serve as a primary key, because two books can share a title while still needing distinct records. You can then ask for records that meet a condition, such as books that are not currently borrowed. The 2026–2028 syllabus includes particular single-table SQL operations, so practise only the syntax and tasks your specification requires.
A useful 25-minute Computer Science practice session
- Minutes 0–4: recall one syllabus term and explain it in ordinary language.
- Minutes 4–9: study a single example, such as converting binary or tracing a loop.
- Minutes 9–15: attempt a different question without notes.
- Minutes 15–21: write or trace a short algorithm, or explain a computer-system scenario.
- Minutes 21–25: check against the specification or marking guidance and redo the first incorrect step.
Common mistakes and immediate fixes
- Memorising definitions without an example: draw or trace what actually happens.
- Reversing binary place values: label powers of two from the right.
- Confusing validation and verification: rules check suitability; verification checks accurate transfer.
- Guessing loop output: write a trace table with the variable state at each step.
- Assuming a working program is fully tested: include normal, abnormal, extreme and boundary data as appropriate.
- Using Python syntax for every Paper 2 coding answer: follow the pseudocode rules outside the scenario question.
- Using the 2029 rules in a 2027 paper: check the examination-year syllabus.
Mini-test: retrieve before you reread
- Convert binary 10011 into denary.
- A two-input AND gate receives 1 and 0. What is the output?
- An algorithm starts with Total = 1 and adds 3, then 5. What is the final Total?
- Is checking that an age is between 0 and 120 validation or verification?
Answers and quick reasons
1. 16 + 2 + 1 = 19. 2. 0; AND requires both inputs to be 1. 3. 1 + 3 + 5 = 9. 4. Validation; it applies a range rule but does not confirm that the entered age is true.
What to do next
Choose one weak area and make a tiny learning card: definition, diagram or algorithm, worked example, new test, common mistake. Revisit it tomorrow. Computer Science becomes faster when you can explain how a system works and design an accurate solution to a new problem, rather than repeating a memorised answer.