CaL: Extending Data Locality to Consider Concurrency for Performance Optimization - 2018

PROJECT TITLE :

CaL: Extending Data Locality to Consider Concurrency for Performance Optimization - 2018

ABSTRACT:

Massive information applications demand a higher memory performance. Information Locality has been the main target of reducing information access delay. Knowledge access concurrency, however, has become prevalent in trendy memory systems in recent times. How to extend existing locality-based performance optimization to contemplate knowledge concurrency becomes a timely issue facing the researchers and practitioners in the sector of computing, especially in the sector of Big Data computing. During this study, we have a tendency to introduce the concept and definition of Concurrency-aware data access Locality (CaL), which, as its name states, extends the concept of locality by considering concurrency. Compared to the standard concept of locality, CaL accurately reflects the combined impact of data access locality and concurrency in fashionable memory systems and is terribly effective for knowledge intensive applications. The worth of CaL will be quantitatively measured directly by performance counters in mainstream commercial processors and is practically feasible. Two theoretical results are presented to reveal the relationships between CaL and existing memory system performance metrics of memory accesses per cycle (APC), average memory access time (AMAT), and memory bandwidth (B). During this approach, we offer a technique to use existing locality-based optimization ways directly or in combination with data concurrency optimizations, to improve the value of CaL and to improve the performance of a memory system. To demonstrate the practical worth of CaL, we have a tendency to conduct four case studies to illustrate the facility of concurrency-aware locality optimization. Compared with the standard locality based mostly optimization, the CaL-aware style has achieved significant performance improvement. It achieved a 3.twelve-fold speedup on K-suggests that, that could be a widely-used information analytic kernel from the big information benchmarks.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here