Parallel Processing

                                          CIS 4100                                                      

Prof. Wasley                                 PAPER 2: PARALLEL PROCESSING                                

SUMMARY

Serial processing is the processor of one processor to perform all processing. The instructions are divided into smaller instruction series and then they are solved each time therefore only one instruction can be executed simultaneously. With parallel computing, we can have Multiple processing units and each unit can run multiple commands simultaneously. Hence different instructions will run on different processors. So, by using parallel computing we can save time, money, computer memory. As of today, technology, the parallel processing takes greater advantage of the multi core processors like intel I-series and AMD Ryzen to improve performance, scalability and efficiency.

BODY

Just like Computer Virtualization, it purpose can be used to make your application easies, allow the application to run on more system compare to one task at a time, improving stability and efficient. On the other hand, In the last three decades, microprocessors evolved by increasing the clock frequency. This approach allows for increased application performance until the processors reach high Gigahertz and result in high heat problem.

To slope problem with over heat problem and to gain better performance, the computing industry moved to parallel processing in 1990s. In 2000s, Multi-core CPUs often have multiple CPU cores on the same die, connected to each other via a shared L2 or L3 cache, a bus on a bus, or a transceiver on a crossbar. All CPU cores on dead connectors share components to communicate with other processors and the rest of the system. These components may include a front-bus interface, a memory controller for communicating with dynamic random-access memory (DRAM), a cache link associated with other processors and a tight link with southbridge and I / O devices, Therefore, the computing industry then able to apply multicore processors with parallel processing in a single processor.

Under low load, parallel processing results in shorter execution time and better for batch processing on a single processor. Parallel processing of tasks and sequential processing tasks is neutral for energy consumption because both modes require the same amount of energy, while parallel processing and processing Different serial rates for maximum saturation workloads as well as for response times. They have been adopted because you can work more when you have parallel processing working at the same time. For the sake of being able to work done faster, the processor can come back to sleep more, so it will use less, or not even power. Serial processors, even if idle, use more power.