Types of Processors

Reduced Instruction Set Computer (RISC)

• A processor architecture that uses a small, simple set of instructions.

• Each instruction is designed to execute in one clock cycle, allowing fast processing.

• Relies on software to perform complex tasks by combining simple instructions rather than complex hardware, making a simpler, faster, and more power-efficient processor.

• Commonly used in smartphones and tablets.

Complete Instruction Set Computer (CISC)

• A processor architecture that uses a large set of complex instructions, each capable of performing multiple low-level operations within a single instruction.

• This reduces the number of instructions per program and simplifies assembly programming. However, CISC instructions take multiple clock cycles to execute, which can reduce efficiency compared to RISC.

• Commonly used in desktop and laptop processors.

Von Neumann Architecture

• A computer architecture model where the processor, memory, and input/output devices share a single communication bus, using a single memory space to store both data and instructions.

• The processor fetches instructions and data sequentially from memory via the Fetch-Decode-Execute cycle.

• This architecture simplifies design but can cause a bottleneck known as the Von Neumann bottleneck, where the processor waits for data/instructions because they share the same bus.

Harvard Architecture

• A computer architecture where the processor uses separate memory and buses for data and instructions.

• This allows the processor to fetch instructions and data simultaneously, improving performance by avoiding the bottleneck seen in Von Neumann architecture.

• Commonly used in embedded systems and microcontrollers where speed and efficiency are crucial.

• However, having separate memory increases hardware complexity and cost.

Superscalar Architecture

• Allows a processor to execute more than one instruction per clock cycle by having multiple execution units.

• It fetches, decodes, and executes multiple instructions in parallel within a single cycle, increasing instruction throughput.

• This improves processor performance by making better use of processor resources and reducing idle time.

• However, performance gains depend on the ability to find independent instructions that can be processed simultaneously.

Embedded System

• A specialised computer designed to perform a dedicated function, often as part of a larger device.

• Typically including a CPU, memory, and connected input/output interfaces to control specific hardware efficiently.

• Optimised for speed, reliability, and low power consumption, rather than general-purpose computing.

• Examples include washing machines, smart thermostats, and automotive control systems.

CPU

• General-purpose processor designed for sequential tasks and system control.

• Has fewer, more complex cores.

• Each core is optimised for high clock speed and task flexibility.

• Ideal for running the operating system, logic-heavy tasks, and varied instruction types.

General Processing Units (GPUs)

• A specialised processor designed to handle graphics rendering and image processing.

• Containing hundreds or thousands of smaller, simpler cores to perform parallel processing, making it highly efficient at handling tasks.

• Also used for general-purpose computing tasks that benefit from parallelism.

• By offloading graphics tasks from the CPU, the GPU improves overall system performance, especially in gaming and multimedia applications.

• Optimised for SIMD (Single Instruction, Multiple Data) execution.

• Ideal for graphics rendering, AI, and scientific simulations where the same operation is applied to large data sets.

Digital Signal Processors (DSPs)

• A specialised microprocessor designed to perform fast mathematical operations on continuous data streams such as sound, images, or sensor signals.

• It is optimised for real-time processing, using features like parallel execution, pipelining, and dedicated hardware multipliers to handle repetitive arithmetic efficiently.

• Improves performance in systems that require constant, rapid computation — by offloading these intensive tasks from the main processor.

Single Instruction, Multiple Data (SIMD)

• A type of parallel processing where the same instruction is applied simultaneously to multiple pieces of data.

• Allowing one control unit to broadcast a single instruction to multiple processing elements, each working on different data values in parallel.

• Highly efficient for data-parallel tasks such as image processing, graphics rendering, and scientific simulations.

• It increases throughput by reducing the number of instructions needed to process large data sets.

Multiple Instruction, Multiple Data (MIMD)

• A type of parallel processing where different instructions are executed simultaneously on different data by multiple processors or cores.

• Each processor operates independently, allowing tasks with varying operations to be processed concurrently.

• Ideal for complex, mixed workloads, such as running multiple programs, simulations, or server-based tasks, improving overall throughput and system performance.

Distributed Computing

• A system where a task is divided across multiple computers or nodes, which work together over a network to complete it.

• Each node performs part of the task simultaneously, often communicating intermediate results to coordinate the overall task.

• This improves speed, scalability, and fault tolerance, as workloads are shared and no single machine handles everything.

• It is commonly used in cloud computing, large-scale simulations, and server farms.

Multicore Systems

• Has multiple processor cores connected into a single processor chip.

• Each core can execute instructions independently and simultaneously, enabling true parallel processing. This improves performance by allowing the system to run multiple programs or threads at the same time.

• More efficient for multitasking and software designed to take advantage of multiple cores.

• Performance depends on how well software can distribute tasks across the cores.

Parallel Systems

• Uses multiple processors or cores working simultaneously to execute different parts of a task at the same time.

• This approach breaks down complex problems into smaller sub-tasks that can be solved simultaneously, significantly reducing overall processing time.

• Effective for applications that can be divided into independent tasks.

• Performance improvements depend on the ability to split tasks and coordinate between processors efficiently.