Super computers, Basic Computer Science

Super Computers:

The specialised demands and requirements of science, industry and military have led to the creation of powerful super computers. For numerically intensive computing (geological exploration, wind tunnel simulations, molecular physics, and space and defence systems), computers capable of performing multiple giga flops (billion floating point operations per second) are being produced. These computers, of which Cray is a well known example, use vector and parallel processing to achieve the necessary number-crunching power. Vector and parallel processing require unusual computer architecture and special programming. For vector processing, the computer architecture works with vector arrays rather than discrete scalar elements, and the numerically intensive portions of a program must be written to use this facility. To achieve full speed and power in vector arrays processing, the super computer must not be interrupted in the middle of its program. which limits the interactive use of super computers. Scientists and engineers have found that they can experiment with more design and research solutions in a short period of time on interactive engineering and graphics workstations.

 Super computers built with a parallel processing architecture are designed to overcome the processing bottleneck caused by using a Von Neumann-type of single instruction stream. The parallel computer architecture must coordinate communication and timing across an array to allow simultaneous computation of subroutines. Although, massively parallel super computers, such as those made by Sequent and Floating Point Systems, are in principle faster and less expensive than the serial super computers, the unconventional programming required to utilise the parallel architectures has been an obstacle to wide wise of these machines.

 It may be stated here that while research continues in the United States and Japan to create faster super computers, some of the techniques employed in super computer architecture are being transferred to microprocessor design. For example, Intel has produced a RISC (Reduced Instruction Set Computing) microprocessor, the 80860, modeled on the Cray super computer. The design uses one million transistors to provide a 64 bit processor, data and instruction caches, integer and floating point math units and a graphics processor. The math units can operate in parallel using a pipe lining system and the data cache. Integrating these functions on a single chip makes it possible to sustain high processor speed over time. The Intel 80860 and similar RISC chips are produced by Motorola and other companies.  

Posted Date: 10/24/2012 4:01:15 AM | Location : United States

Related Discussions:- Super computers, Assignment Help, Ask Question on Super computers, Get Answer, Expert's Help, Super computers Discussions

Write discussion on Super computers
Your posts are moderated
Related Questions
The Widening Web The idea, the whole world is agog now had accidental origins. Here is how it all began. In early 1969, the U.S. Defence department financed a network fo

I need to create mulitple queues assignment that run concurrently. It is like a queue simulation with  number of checkouts with a random probability of clients, joining any one

Data communication as a need developed in the 1960s with the interconnection of peripheral devices to mainframe computers. Within the immediate vicinity of the mainframe computer t

The difference between kernel mode and user mode gives a rudimentary form of security in the following manner. Convinced instructions could be executed only when the CPU is in kern

Input/output (I/o) unit: This unit provides the interface between the computer and the computer peripherals. A computer peripheral is any unit, which is attached to, but is no

Question 1 Explain the important concepts of modern CPU CISC vs. RISC CPUs Circuit Size and Die Size Processor Speed Processor Cooling System Clocks Architect

Data Output : Processed data is no use to you if you cannot use it. Once data has been processed, you will either: (1) Send it as a data file to another system, e.g. write a fi

Specifying Search Problems In our agent terms, a problem to be solved is a particular task where the agent starts with the environment in a given state and acts upon the enviro

Artificial Intelligence Agents In the earlier teach, we discussed what we will be talking about in Artificial Intelligence and why those tasks are important. This lecture is all

construct a tree where preorder is ABCDFGE