2 additional insight writings

Instructions: Add additional insight opinions or challenge opinions and you can visit a couple of the web sites contributed and share your opinion of these sites.  Minimum of 150 words for each. 

Write additional insight, up to 150 words 1) Parallel computing uses two or more processors rather than single processors, to perform computations and computer programming functions. The single processing concept of the “old days” were considered serial computing processing. Parallel computing allows a computer or system to process several different task at once. In serial computing, processing took longer due to waiting to process jobs (Stout, 2017, para. 6). Also, most computer science departments at universities and colleges begin the teaching curriculum with serial or sequential programming methods. In today’s world of programming, the vast majority of applications run using multiple cores (parallel processing); it makes since that the curriculums should change to accommodate this era (Kirkpatrick, 2017, pgs. 17-19). Parallel Computing is very significant to high performance computing systems (Craus, Birlescu, and Agop, 2016). The Graphic Processing Unit (GPU) has become a standard in parallel processing due to its low cost and massive processing footprint (Navarro, 2014, pg. 285). The introduction of the GPU to computer processing, with it processing power, has significantly enhanced the Central Processing Unit (CPU), which has always been known as the brains of the computer. However, even with the advent of the GPU, there is still and issue in parallel computing with bottleneck processing. Resistive switching memory, called RRAM, has helped with some of these bottlenecks, but can be improved by a proposed parallel architecture that is the result of pattern recognition (Jiang, et. al., 2017, para. 1). Although developing of a more streamlined way of parallel processing with bottlenecks, has not been totally introduced, there has been advances over the years compared to serial computing. To this point, discussion of parallel processing has been hardware-based. There are also software-based parallel computation. One of these techniques is processing using “R”, which is a statistical programming language. This language is used to nest calculations and speed processing because of built in libraries that already have the capability for parallel processing (Mount, 2016, pg. 1). The R program is open source software that can run on Windows Linux/UNIX, and the Mac OS. Together, hardware and software has been the solution to providing parallel computing, which is known is very pertinent to High Performance Computing.

Write additional insight, up to 150 words

2) Parallel computing is best understood when compared to serial computing, which is how most computers have been programmed and which most people are familiar. In serial computing a problem is programmed into a discrete set of steps that will be executed in sequence. A single CPU executes each step one after another until the program ends. It's important to understand that only one step can be executed at any given time on the processor (Blaise, 2018, para.1). Eve.n though modern computers have extremely fast processors they still have limits with resources, memory and bus speed, all of which must be taken into account Parallel computing uses multiple processors or multiple networked computers to solve a problem much faster than it could be solved serially (Stout, 2018, para.1). Each CPU gets a piece of the overall task and works to complete it (Krastev, n.d., p.6). The more processing units available, the faster the task can be completed. Some form of coordination process is employed to efficiently manage the overall task and monitor the progress.  An example may be helpful. Let's say we wanted to plot an accurate graph of a quadratic equation. In serial computing each value of “x”, “y”, and other variables would have to be computed one at a time, one after another. For a small graph we (or a single processor) could do this fairly rapidly but as the graph got larger we'd need more time to compute each new value. Depending on the final size or accuracy needed this could take a significant amount of time. Using parallel computing we could give each processor or separate computer a different value to compute each point. In this scenario the problem could be finished many times faster than on a single computer or processor (Eijkhout, 2011, p.7).  Perhaps the most well known use of parallel programming presently is the use of GPUs in Bitcoin and cryptocurrency mining. The ability of GPUs to run tens of thousands of threads simultaneously and their relatively low price when compared to other solutions make them attractive to Bitcoin miners. Number of Pages: 1 Page Page Line Spacing: Double spaced (Default) Number of Slides : No slides needed Deadline: 5 days Academic Level: College Paper Format: APA