what about single core system and multi core system and its synthesis report analysis
Answers
Answer:
We argue that in order deliver real-time event processing in presence of continuous proliferation of data and bandwidth, it is inevitable to expand the research horizon beyond the conventional sequential matching algorithms and adopt other key enabling technologies such as multi-core architectures. There is a mounting evidence that the number of cores is predicted to grow exponentially in the coming years [14], [18], and that multi-cores are successfully being used to accelerate many data management applications [15]. Therefore, the next generation of real-time event processing must exploit parallel hardware such as multicore architecture to sustain the expected growth of demands. ...
... First, the database indexing solves the reverse problem: in the database context, querying means finding the relevant tuples (events) for a given query (subscription), but in event processing context, matching (through indexing) means finding the relevant subscriptions (queries) for a given event (tuple). Second, publish/subscribe matching algorithms overlook both parallel event matching and event stream re-ordering, which is central in exploiting the exponential growth trend in the number of cores of modern hardware [14], [18]. ...
... This problem is further amplified as the number of parallel tree traversal are increased, resulting in a higher number of cache-misses. In addition, BE-Tree fails to algorithmically exploit the large number of cores and large shared caches prevailing in multi-core architectures [18]. In this work, we present novel adaptive parallel compressed event matching (A-PCM) and online stream re-ordering (OSR) algorithms that exploit all these new hardware properties. ...