Evolution of Computer Technology - Essay Example

Evolution of Computer Technology Computer technology has evolved over the last 25 years to provide the near perfect conditions that are experienced in the current age. Much of the evolution that has seen the better performance of computers involves the improvement of the speed and performance of computers. Since computers process large amounts of data, most of the improvement is focused on processing speed, hence improving the performance of the computer. To analyze this evolution, the main factors that should be analyzed are interrelated and include Reduced Instruction Set Computers (RISC), pipelining, virtual and cache memory. The use of RISC overlaps many concepts in computer technology, with pipelining being one of the main characteristics of RISC (Cocke and Markstein, 2000). Pipelining typically refers to the improvement of the speed of processing instructions in a computer processor. In this concept, multiple instructions overlap in processing so that more instructions are processed. Instead of the computer processing one instruction at a time, many instructions are processed simultaneously. The processing of instructions is done in machine cycles, where one instruction is processed in one stage and moved to the next while another instruction is processed in the first stage. The main stages of pipelining include fetching the instruction, decoding, execution, memory access and write back (Cocke and Markstein, 2000). Similarly, superscalar processing improves the notion of pipelining in RISC. Earliest evolutions of RISC processors aimed to execute at the rate of one instruction per cycle, but with the introduction of superscalar processors, this has been improved to more than one instruction per cycle. For the effective performance of computer technology, the memory that affects the speed needs to be effectively managed. According to Cocke and Markstein (2000), memory system management refers to the allocation and security of portions of computer memory to different programs. Memory system management helps in multitasking in the computer since it utilizes the concepts of pipelining and superscalar processing to ensure that computer programs are effectively managed. Memory system management is usually covered in four main concepts, with the main one being relocation. Cocke and Markstein (2000) indicate that virtual memory management should account for the storage of different parts of memory in different places, and also provide for the recalling of this memory. When a program is swapped in a pipeline, it is not placed in the original memory, which means that effective memory management should tell where the program is stored for easy recall. The second concept is that of protection, where the different processes should be protected from each others’ references to avoid corruption (Carter, 2001). However, the different processes sometimes need to share memory, so memory system management should also be able to account for the sharing of information. Shared memory therefore forms part of inter-process memory, which effectively increases processing speed. Finally, memory system management should have effective methods of allocation of memory to different processes and programs in the computer. Logical allocations of the different modules in a program ensure that programs that need to share references are allocated effectively and others separated in other parts of memory. An example of effective memory allocation methods in memory system management is segmentation, where the memory is separated according to modules. One other concept that is used to increase the efficiency of processors is the use of caching. The use of cache memory is important since it is the bridge between the processor and the memory in a computer, so temporary instructions are usually stored in cache memory (Carter, 2001). Cache memory stores part or all of a program when it is being processed and also stores the data required by the processor for different factions. Some of the strategies used in caching to increase efficiency include replacement strategies in associative caches, where information that is not needed in the cache is replaced by other information to speed up processing. The need for replacement techniques is driven by the relatively small size of cache memory; it cannot be compared to the size of the memory or the processor. Another strategy used in cache memory is write strategy, which usually includes write-through and write-back strategies. Write-through strategies are more memory consistent but increase the traffic of data and information in cache memory (Carter, 2001). Conversely, write-back strategies are less intensive on the memory but require more management since they change constantly. Memory management requires the computer to allocate memory to different programs when needed (Carter, 2001). However, physical memory in the computer is usually limited, hence the need of back-up memory referred to as virtual memory. Virtual memory usually refers to a replacement set of memory addresses that back-ups physical memory addresses. These virtual memory addresses in a computer are used to store physical addresses in a way that can be recalled when a program needs to use the address. Since physical memory is limited, a computer will store some part of memory in a virtual paging file, and when programs need to be executed, the virtual addresses are copied to the main memory and executed by the computer (Carter, 2001). Virtual memory is mainly used by the computer to increase the address space that can be used to execute programs. The evolution of the factors discussed above can be traced back to the need of improved processing and the reduction of complexity in computer instructions. RISC technology evolved in IBM technologies with the need to address how instructions are addressed in programs (Cocke and Markstein, 2000). Since many instructions are processed in a computer, researchers viewed the most frequently used instructions and decided to give them priority in a computer. This evolution included the use of pipelining to address the multiple processing of different instructions. Actually, the evolution of RISC can be tied together with the evolution of virtual memory, caching and pipelining since this concepts support each other to ensure the efficient processing of instructions. The evolution of virtual memory started in early 1960s, where computers were developed to translate virtual addresses into real addresses. Intel first offered virtual and cache memory as a way of assigning memory in their processors, a concept that later developed and included in ost computers for everyday use. The evolution of pipelining can be traced down to the use of CISC, but in the RISC was introduced to simplify ISA and improve pipelining. At this time, Stanford introduced architecture that improved their compilers while architecture that did not require much compiling was introduced at Berkeley. In the 90s, pipelining was improved with the use of simplification and scheduling to enable re-ordering of instructions and introduction of more transistors. This also introduced the use of parallel instructions to improve pipelining in the processor. From the above discussion, it is evident that RISC technology is inherently tied up with the use of pipelining, virtual memory and cache memory. The choice of discussion chosen above reflects the way these terms interact to give the best performance of the computer. The evolution of RISC and its effectiveness is shown, and followed up by the concept of pipelining which helps the processor work on different instructions at the same time. Pipelining as a concept can be related to the use of virtual memory which technically increases the limit of memory in the computer and improves processing speed. In the same way, the use of caching in the processor helps in reducing the machine cycle time taken in processing instructions since the data between programs are stored in easy to reach locations. From this discussion, it can be seen that the evolution of the factors discussed is especially important in ensuring that the computer performs effectively. References Cocke, J. and Markstein, C. V. (2000). The evolution of RISC technology at IBM. IBM Journal of Research and Development, Vol. 44(2). Pp.48-55. Carter, N. (2001). Schaum’s Outline of Theory and Problems of Computer Architecture. New York: McGraw Hill Publishers. Read More