Parallel Processing Shared Memory with Snoopy Cache

Provided by: IRD India
Topic: Hardware
Format: PDF
Implementation of parallel processing with snoopy cache is an idea to enhance the performance of computing in real-time. When the authors consider uniprocessor the execution will take sequential. In uniprocessor they may not face problems. When they come to multi-core processors in parallel processing the program is divided into slices and each slice may be independent or dependent. If it is independent then no problem, but dependent case is different. There may be variables which may be used anywhere in program. Such variables or data is passed to a shared memory. That memory should be tracked or notified to the CPUs in execution process. There are some problems in normal process.

Find By Topic