Multi-Core Processing and Parallelism Fundamentals Quiz Quiz

Explore core concepts of multi-core processing and parallelism with this quiz, designed to help you identify key features, benefits, and common terminology in parallel computing and CPU architectures. Perfect for those looking to understand how modern processors handle multiple tasks efficiently.

1. Multiple Cores in Modern CPUs

   What is the primary advantage of having multiple cores in a CPU when running several applications at once?

   1. Increased cost of the computer system
   2. Slower overall performance due to complexity
   3. Reduced memory usage in the system
   4. Improved multitasking by processing tasks simultaneously

   Explanation: Multiple cores in a CPU allow different tasks to be processed at the same time, significantly improving multitasking and overall system responsiveness. Increased cost is often a result of multi-core CPUs, but it is not an advantage. Slower performance is incorrect; multiple cores are designed for speed. Memory usage mainly depends on running programs and not directly on the number of cores.

2. Definition of Parallelism

   Which of the following best describes parallelism in computing?

   1. Performing multiple operations at the same time
   2. Using less electricity for one process
   3. Running tasks in a strict sequence
   4. Storing more data in cache memory

   Explanation: Parallelism specifically means executing multiple operations or tasks at the same time, often using multiple cores or processors. Running tasks in sequence is called serial or sequential processing. Using less electricity is related to power efficiency, not parallelism. Storing more data is a function of memory, not processing.

3. Task Scheduling Example

   If a computer has a quad-core CPU and needs to sort several large lists, how can parallelism optimize this task?

   1. By writing data twice in memory
   2. By increasing the screen resolution
   3. By dividing the lists among the four cores to sort simultaneously
   4. By sorting the lists one after another on a single core

   Explanation: Parallelizing the sorting task by distributing lists among multiple cores speeds up overall completion. Sorting lists one by one would ignore the benefits of multiple cores. Writing data twice in memory is unrelated to this scenario. Screen resolution changes have no impact on processing tasks like sorting.

4. Understanding Core vs. Thread

   Which statement correctly differentiates a CPU core from a thread?

   1. Cores are only found in graphics cards
   2. A core is a physical unit of processing, while a thread is a virtual sequence of instructions
   3. Threads consume more power than cores
   4. A thread is always faster than a core

   Explanation: A core is the actual hardware capable of independently executing instructions, while a thread represents a sequence of tasks scheduled for execution. Threads are not inherently faster than cores; in fact, performance depends on implementation. Cores exist in many types of processors, not just graphics cards. Power consumption varies, but threads themselves are not physical units.

5. Shared Memory in Multi-Core Systems

   Which of the following is typically shared among all cores in a multi-core processor?

   1. Separate power supplies
   2. Main system memory (RAM)
   3. Individual clock generators
   4. Operating systems for each core

   Explanation: Cores in a multi-core processor usually access the same main system memory, allowing efficient data exchange. Clock generators may synchronize the cores but are not individual for each. There is generally one operating system overseeing all cores, not one per core. Cores receive power from the same supply rather than separate ones.

6. Embarrassingly Parallel Problems

   What does the term 'embarrassingly parallel' describe in computing?

   1. Embarrassing coding mistakes in programs
   2. Operations that can only be run on a single core
   3. Tasks that are hard to parallelize due to dependencies
   4. Problems that can be split into independent tasks easily

   Explanation: An 'embarrassingly parallel' problem can be divided into many independent subtasks that do not rely on each other, simplifying parallel execution. If tasks have strong dependencies, they are not considered embarrassingly parallel. Coding mistakes are unrelated. Operations that require a single core are the opposite of parallel problems.

7. Load Balancing

   Why is load balancing important in parallel processing environments?

   1. To increase the system temperature deliberately
   2. To make all applications run in sequence
   3. To reduce the number of tasks to a minimum
   4. To ensure all processing units work equally and efficiently

   Explanation: Load balancing distributes work evenly across all cores or processing units, maximizing resource usage and minimizing delays. Running applications in sequence does not leverage parallelism. Reducing task count isn't always beneficial if it means underutilizing processors. Intentionally increasing temperature has no functional advantage.

8. Term for Executing Same Instruction on Multiple Data Elements

   What is the term for a process where a single instruction operates simultaneously on multiple data elements?

   1. SIMD (Single Instruction Multiple Data)
   2. RAMD (Random Access Multiple Data)
   3. FIMM (Fast Instruction Multi Mode)
   4. MISD (Multiple Instruction Single Data)

   Explanation: SIMD allows one instruction to process several data points in parallel and is widely used in tasks like graphics processing. MISD is a rare architecture where different instructions act on the same data. FIMM and RAMD are incorrect and do not represent standard computing models.

9. Bottleneck Causes in Parallel Processing

   What is a common cause of bottlenecks in parallel processing when multiple cores try to access shared resources?

   1. Improving cache efficiency
   2. Contention for shared memory or data
   3. Running only one application at a time
   4. Having more than one screen attached

   Explanation: Bottlenecks commonly occur when several cores compete to access the same memory or data, causing delays and reduced performance. The number of screens is unrelated to CPU resource management. Improving cache efficiency helps decrease, not increase, bottlenecks. Running one application rarely causes contention.

10. Programs Designed for Multiple Cores

    What term describes software written to take advantage of multiple cores by dividing its tasks among them?

    1. Parallelized
    2. Serialized
    3. Fragmented
    4. Compressed

    Explanation: Parallelized software distributes tasks to be executed concurrently across multiple cores, improving performance. Serialized programs do the opposite by running tasks one after another. Compression relates to reducing data size, not core utilization. Fragmentation deals with data storage, not program execution.