<![CDATA[The growth of technology will impact to the growth of data beyond limits of database management tools. One of the system is Information Management System for Hospital it’s a high complexity problem solving method related to the load of data. Parallel computing is one of the technique that been used in HPC. The focus of this research will be emphasized to the design of parallel computing using message – passing model as a search system process in Information Management System for Hospital to find patient data. The design of parallel computing will be done in a way to shared computing data to a number of CPU (Master and Slave), the parallel computing configuration used on CPU (Master and Slave) will be using some stage FOSTER method that are partition, communication, aglomerasi and mapping. The test will be done computing data processing time between sequential and parallel time. Parallel computing design will be tested using speedup and efficiency calculation. The result of designing and testing parallel computing using message-passing model proved the patient data processing speed using parallel program is capable to overcome 1 CPU using sequential network topologi. On speed up method , the test indicate an increase on data transfer speed up using 3 parallel computing CPU. While using efficiency testing method, the efficiency point reached its peak when using 2 and 3 CPU. The Occurrence of decrease on speedup and efficiency point were caused by the minimal amount of data if handled by 7 parallel computing CPU. The Conclusion for this method is, the increase amount of CPU involved in the data processing using parallel computing is not proportional to the amount of time to processing data itself. It happened because every data processing task in terms of the amount of data that’s handled have an ideal amount of CPU limits to do the task itself.]]>
Copyrights © 2018
