What is the fundamental difference between concurrent and parallel processing?

Concurrent processing involves interleaving multiple tasks on a single processor core using time-slicing. Parallel processing involves executing multiple tasks simultaneously across multiple different processor cores.

How does a single-core processor achieve the appearance of running multiple programs at once?

It uses time-slicing, where the CPU allocates a tiny fraction of time to one process before rapidly swapping to the next. Because this happens in microseconds, it appears simultaneous to the user.

Why can't all problems be solved using concurrent processing?

Many problems have 'dependencies' where one task requires the output of another to begin. These tasks are inherently sequential and cannot be interleaved or run in parallel effectively.

What is 'throughput' in the context of concurrent processing?

Throughput refers to the total amount of work or the number of tasks completed within a specific timeframe. Concurrency increases throughput by allowing multiple tasks to make progress rather than finishing one at a time.

What is a 'dependency' in task analysis?

A dependency exists when a task relies on the completion or data of a previous task. Identifying these is crucial because dependent tasks cannot be tackled concurrently.

What is the main disadvantage of concurrent processing regarding CPU time?

The main disadvantage is the overhead caused by context switching. The CPU must spend time saving and loading the states of different tasks, which is time not spent on actual computation.

When is parallel processing preferred over concurrent processing?

Parallel processing is preferred for computationally intensive tasks that can be split into independent parts, as it utilizes multiple cores to reduce the total wall-clock time required for completion.

What error occurs if a developer assumes concurrency will always speed up a single task?

This is a misconception; concurrency improves the efficiency of managing multiple tasks but can actually slow down a single task due to the added management and switching overhead.

Define 'Time Slicing'.

Time slicing is the technique of assigning a specific, very short period of CPU time to each process in a multi-tasking system to ensure all processes get a turn to execute.

How does pipelining relate to concurrency?

Pipelining is a form of concurrency where different stages of several tasks are performed at once. It allows a new task to begin its first stage while the previous task moves to its second stage.

Concurrency in Problem Solving | OCR AS-Level Computer Science

6. Elements of Computational Thinking

Concurrency in Problem Solving

Summary

Concurrency in problem solving is a computational strategy where multiple tasks are managed by interleaving their execution on a single processor. By utilizing time-slicing and identifying independent sub-tasks, systems can provide the illusion of simultaneous progress, maximizing resource efficiency and program throughput.

1. Definition & Core Concepts

Concurrent Computing is the execution of multiple tasks where the start and end times overlap, but they do not necessarily run at the exact same instant. It involves a single processor switching between tasks rapidly, giving each a small 'time slice' of execution.

The primary goal of concurrency is responsiveness and throughput, allowing a system to handle multiple requests or background processes without waiting for a single long-running task to complete.

Time Slicing is the mechanism where the Operating System allocates a specific number of microseconds to a process before preempting it to allow another process to run, creating the appearance of parallel execution to the human observer.

Diagram showing Task A and Task B interleaving over a single timeline, illustrating how concurrency switches between processes.

2. Underlying Principles

The foundation of concurrency lies in Task Independence. For tasks to be handled concurrently, they must be able to make progress without requiring the immediate output of another currently running task.

Interleaving is the logical arrangement of instruction sequences from different tasks into a single stream of execution. This allows the CPU to remain active even if one task is waiting for an external event, such as user input or data retrieval.

Non-determinism is a key characteristic; because the exact timing of task switches depends on the scheduler, the precise order of execution steps across different tasks may vary between runs.

3. Methods & Techniques

Dependency Analysis is the first step in concurrent problem solving. Developers must identify which parts of a problem rely on previous results (sequential) and which can operate autonomously (concurrent).

Pipelining is a technique where different stages of multiple tasks are processed concurrently. While one task is in the 'output' stage, the next can be in the 'processing' stage, and a third in the 'input' stage.

Context Switching is the procedural overhead where the state of a process is saved so it can be resumed later. Efficient concurrency requires minimizing this overhead to ensure the benefits of multitasking outweigh the cost of switching.

4. Key Distinctions

5. Exam Strategy & Tips

6. Common Pitfalls & Misconceptions

Concurrency in Problem Solving

Summary

1. Definition & Core Concepts

Diagram showing Task A and Task B interleaving over a single timeline, illustrating how concurrency switches between processes.

2. Underlying Principles

Non-determinism is a key characteristic; because the exact timing of task switches depends on the scheduler, the precise order of execution steps across different tasks may vary between runs.

3. Methods & Techniques

4. Key Distinctions

It is vital to distinguish between Concurrency and Parallelism. Concurrency is about dealing with many things at once (structure), while parallelism is about doing many things at once (execution).

Feature	Concurrent Processing	Parallel Processing
Hardware	Single or Multiple Cores	Multiple Cores/Processors Required
Execution	Interleaved (Sequential at micro-level)	Simultaneous (Truly at the same time)
Goal	Responsiveness and Throughput	Computational Speed and Performance
Complexity	High (Managing state/switches)	Very High (Managing synchronization/data)

5. Exam Strategy & Tips

Identify Independence: When asked to justify concurrency, always look for sub-tasks that do not share immediate data dependencies. If Task B needs Task A's result to start, they cannot be concurrent.
Use Precise Terminology: Avoid saying tasks happen 'at the same time' for concurrency; instead, use terms like 'interleaved,' 'time-sliced,' or 'making progress in the same timeframe.'
Explain the 'Why': If a scenario involves a user interface and a background calculation, explain that concurrency is necessary to keep the UI responsive while the calculation proceeds.
Check for Overhead: In evaluation questions, mention that if tasks are too small, the time spent switching between them (overhead) might exceed the time spent actually processing them.

6. Common Pitfalls & Misconceptions

A common misconception is that concurrency always makes a single program run faster. In reality, concurrency often makes the system more efficient by handling multiple programs, but a single task might take slightly longer due to context switching.

Ignoring Resource Contention is a frequent error. Even if tasks are logically independent, they may compete for the same hardware resources (like a hard drive or network card), which can create bottlenecks that negate the benefits of concurrency.