Why Parallel Programming?

In spite of the fact that multiprocessors have numerous favorable position it additionally have some detriment like complex in structure when contrasted with uni-processor framework.

4. Applications where various of machines can be doled out for each to do a task e.g every processing a single file

“Moore’s law is the observation that, over the history of computing hardware, the number of transistors in a dense integrated circuit has doubled approximately every

Special hardware can differentiate the multiple processors, or the software can be written to allow only one boss and multiple workers. For instance,

In the past processors required only a boost in their clock speed to have an improvement in efficiency, but as technology evolved, companies like Intel, a major processor manufacturer, came up with

The evolution of the processor has come all the way from the Intel 4004 to the processors of today and one would expect Moore’s law to continue to allow exponentially more powerful processors from here on out. Moore’s law will continue, but in a way of overall speed instead of transistors. Modern processors being so fast they produce too much heat to effectively cool, leading to CPU developers needing to design new cooling systems or a new way of making transistors entirely. Another hurdle for CPU developers is that as the transistors become smaller and smaller, quantum tunneling takes place, and a semiconductor becomes just a conductor, allowing power all of the time. A possible solution being quantum computing.Quantum computing involves the bits used in storing data being individual atoms. It’s theorized that such computing techniques could be used to simulate complex chemical processes and increase the accuracy of atomic clocks. We’ve seen where CPUs have come from, we’ve seen where they are, now we have the chance to see where they will go, and the extraordinary heights they will

Computing speed and storage capacity has been doubling about 18 months for the past 30 years. “This adherence to Moore’s Law has produced astonishing advances in communication, genetics, medicine, manufacturing, entertainment, transportation and even warfare” (AASCU 2012).

Starting from the basics the CPU (central processing unit) also known as a processor is a microprocessor contains multiple cores that process itself. In a system you would find many CPU’s and if you open the CPU case up then you will find that the socket(processor) is inside and if you zoom in again you will also find that a processor is made up of core’s. There are two types of processors a single core processor and a multiple processor the difference between them is that the single-core processor can complete one task at a time while a multiple core processor can run two programs. A popular and successful solution of multi-core from Intel meant that you could perform 2 tasks at the same time

The challenges for OS structures depend on the diversity of hardware like number of cores, memory hierarchy, IO configuration, instruction sets and interconnects.

Parallel processing computers have many processors working in parallel; the difficulty is to make sure the problem is synchronized properly and the programs are organized to take advantage. Cray Computer's supercomputer XT3 is a scalable, parallel design which can be scaled up to very huge. In a single cabinet, it can hold 96 AMD Opteron processors, which can operate at 998 GFlops per cabinet using dual-core processors. To support this it has a 64K instruction cache, a 64K data cache, and a 1 MB L2 cache per processor, and up to 8 GB ECC RAM per cabinet; the RAM operates at 6.4 GB/sec per processor. The real secret of the XT3's power is the routing and communications ASIC for each processor. This not ties together the processors within a cabinet,

When one executes something synchronously, one has to wait for it to be finished before moving on to another task. But if one executes something asynchronously, one can go on to another job before it finishes. Asynchronous pro-gramming is a means of parallel programming. A unit of task can be executed separately using multi-thread or some another method in asynchronos programming. A thread is a series of commands or block of code that exists as a unit of work.

Computers today have developed from running single program capability and running run one program at a time to having the ability to run multiple programs at the same time. They are also able to use multiple threads to provide more than one task to be run at the same time. Processes were created to help manage the execution of the programs. A process is defined as a unit of work in a modern time-sharing system during the execution of a program. There are five states that a process may be in new, running, waiting, ready, and terminated. Only one process can be running on a processor and the other processes are in a ready and waiting state.

Software industry is examining the agile software development methodologies by adopting them to few of its software projects. Industry, particularly those following the eXtreme Programming (XP) methodology, has popularized the use of pair programming. Developers often require several days to become familiar and comfortable with the dynamics and practices of pair programming when transitioning from solo programming. Extended pair programming sessions can be difficult for developers because pairing can be mentally exhausting due to the rapid pace a pair can work at and the constant focus on the task at hand. Often software engineers feel pair programming should be saved for specification,

Moreover, the customers will never feel satisfaction with buying the same or few types of microprocessors. In order to satisfy various demands, Intel always spend huge amount and put a lot of efforts into their R&D department which makes Intel is able to introduce new products before its competitors. Consequently, Intel can provide the choices to the demand from high-end product with high price and the low-end demand with lower price.

decided to change the traditional process to increase productivity – 1> Increased parallelization of design tasks, 2> Elimination of