A processor is the most important integrated circuitry (IC) in computer. It is used for interpreting most computer commands. It is an electrical component that performs operation on external data sources such as memory and data stream. Typically taking the form of microprocessors, it can be implemented on integrated circuit chips.
In this article on ‘what is processor’, we will discuss about this circuit, its components, types and examples.
What is Processor in Computer?
Processor or Central Processing Unit (CPU) refers to logic circuitry that responds to and processes basic instructions that drive computers. This integrated electronic circuit performs calculations that run computers. These circuits are found in electronic devices. They receive input in the form of program instructions and execute calculations for providing with which the user will interact.
For any operation to perform on a computer, the processor is required to interpret the operating system. A processor consists of arithmetical logic and a control unit (CU) that measures capability in terms of:
- Maximum number of bits/instructions
- Relative clock speed
- The ability of processing instruction at a given time
Components of a Processor
A processor has four components, including a floating point unit (FPU), an arithmetic logic unit (ALU), registers, and cache memories.
1. Arithmetic Logic Unit (ALU)
ALU is the main component in a processor that performs various arithmetic and logic operations. It is an integrated circuit within the CPU/GPU, due to which it is also known as an integer unit (IU). This is the last component that performs calculations in the processor.
2. Floating-Point Unit (FPU)
It is part of the computer system that is used for carrying out operations on floating-point numbers. These operations include square root, multiplication, division, subtraction, and addition. It can perform transcendental functions such as trigonometric and exponential functions; however, it may not be accurate.
3. Registers
Registers are types of computer memory that accept, transfer, and store data as well as instructions that are being used. It instructs ALU about the processes that must be carried out and stores the result of these operations.
4. Cache
Cache is the smaller yet faster memory that is located close to the core of the processor. This memory stores the copy of data from the frequently used main locations. There are three levels of cache: L1, L2 and L3 cache. L1 is the primary chip which is embedded on the processor chip.
Since it is small, it has limited storage. L2 cache is the secondary cache that is either embedded on a processor chip or a separate chip with high-speed bus that connects it to the CPU. Also known as processor cache, L3 is a specialized backup memory for L1 and L2. It boosts the performance of L1 and L2.
Types of Processors
Let us now discuss the different types of processors that are available at present.
1. Application-Specific Instruction Set Processor (ASIP)
It is a component used in the system-on-a-chip design. The instruction set of ASIP is customized to benefit specific applications. For certain ASIPs, this instruction set is configurable. ASIP can be used as an alternative to hardware accelerators for video coding or baseband signal processing.
Explore networking and hardware courses
2. Processor based on Flynn’s classification
According to Flynn’s taxonomy, processors can be classified based on concurrent instructions and data streams that are available in architecture. Let us now discuss each of the above one by one.
2.1 SISD (Single Instruction Single Data)
It is a computer architecture in which single uni-core processor executes single instruction stream. This is done for operating on data that is stored in single memory. SISD can have (as per Flynn) concurrent processing characteristics.
Here, instructions are sent to control unit from memory module. Then, they are decoded and sent to processing unit that processes data retrieved from memory module and then sends it back. Traditional uniprocessor machines such as PCs, old mainframes, pipelined and superscalar processors are examples of SSID.
2.2 Single Instruction Multiple Data (SIMD)
It is a type of computer that comes with multiple processing elements. It simultaneously performs same operation on multiple data points as well as parallel computations on only a single instruction at a given time. SIMD may be a part of the hardware design and is directly accessible through instruction set architecture (ISA). These machines do not exploit concurrency.
2.3 Multiple instruction Single Data (MISD)
It is a type of parallel computing architecture where multiple functional units perform different operations on the same data. Every CU here handles one instruction stream and processes it through corresponding processing elements. It has an architecture that is used for fault tolerance. MISD organisation computers are used rarely. Space Shuttle flight control computer is an example of MISD.
2.4 Multiple Instruction Multiple Data (MIMD)
This refers to a technique used for achieving parallelism. Machines that have MIMD have several processors that function independently and asynchronously. Multiple autonomous processors execute, at any time, execute different instructions on different data pieces.
These machines can be either of shared or distributed memory categories based on how MIMD processors are accessing memory. Shared memory may be bus-based, hierarchical or extended types. Distributed memory may be of hypercube or mesh types.
3. Processor Based on Number of Cores
Following are the different types of processors based on number of cores:
3.1 Single core
A single core microprocessor has single core in its die. It performs ‘fetch-decode-execute cycle’ once per clock cycle since it runs only on a single thread. These processors have been less in demand due to lesser processing power. Their slow speed has made multi-core systems more popular.
3.2 Multi-core
Multi core processors are microprocessors on a single integrated unit having two or more cores. Each core reads and executes the program instructions. Here, a single processor can simultaneously run instructions on separate core. Due to this, the overall speed for programs supporting multithreading and parallel computing techniques increases.
3.3 Hyper-Threading
It is a technology that is used in the Intel microprocessors. This technology allows a single microprocessor to act as two processors for the operating system as well as the application using it. Through hyper threading, processor resources are more efficiently used which allows multiple threads to run on each core.
4. Special processors
Following are the different types of special processors:
4.1 Graphics processing unit (GPU)
It is a specialized electronic circuit that manipulates and alters memory for accelerating the creation of images in frame buffers that are intended for output to display devices. They can efficiently manipulate image processing and computer graphics.
Due to their highly parallel structure, they are more efficient than general-purpose CPUs for those algorithms where processing of large data blocks is performed in parallel. GPU may be embedded on motherboards or on video cards.
4.2 Physics Processing Unit (PPU)
Also known as the Physics Acceleration card, it is a dedicated microprocessor that handles physics calculations, unlike GPU. It is used specifically for the physics engine of video games. This microprocessor helps in offloading time-consuming tasks for the Central Processing Unit of computer. It provides physics simulation data and communicates this data to the CPU. These are used in high-performance computers.
4.3 Digital Signal Processor (DSP)
It is another specialized microprocessor having an architecture optimized for operational needs of digital signal processing. This measures, compresses and filters continuous real-world analog signals. They are more power efficient, due to which they can be used in portable electronic data.
These processors fetch multiple instructions and data at the same time. DSPs are cost-effective since they are cheaper yet provide better performance and lower latency. They do not have any requirements for specialized cooling or larger batteries.
4.4 Network processor
It is a special-purpose hardware device that is programmable. These are low-cost and flexible, like RISC processors, and scalable and fast as ASIC chips. Such processors are used for designing networking applications.
They have characteristics that are similar to general-purpose CPU used in different types of equipment and products. Firewalls, routers, switches and network security devices use network processors.
Explore free networking courses
4.5 Front end processor
These are smaller computers that connect networks to host computers. Here, data is transferred between front-end processor and host computer through high-speed parallel interfaces.
They offload the host computer from the task of managing peripheral devices, packet assembly and disassembly as well as error detection and correction. These processors communicate with peripheral devices using serial interfaces via communication networks.
CONCLUSION
Hope that this article on ‘what is processor’ has been able to help you understand what processors are and their types in detail. You can choose among the above-mentioned processors based on your speed, processing, and storage requirements.