Within this open, multi-partner cooperative working group, the first phase of benchmark development will include graph analytics, memory caching, and media serving.
This application consumes big-data data sets (e.g. social media output) and uses graph algorithms to find connectivity and common node qualities. An example of graph analytics is page rank which derives website popularity from social data; it’s also used for applications such as Facebook and Twitter. For this benchmark workload, EEMBC will implement a standardized implementation of page rank. This benchmark could report metrics such as time to completion (or convergence) of the graph.
This application enables real-time streaming for on-demand access using large server clusters to packetize and transmit media files. The data center services use large server clusters to adjust and deliver quality based on various pre-encoded formats and bit-rates to suit a wide range of client devices. Industry examples include NetFlix, YouTube, and Pandora.
Memory Caching Analysis
Caching is used in data centers to optimize performance and energy usage, helping to minimize latency and allowing data centers to meet the quality of services requirements of all big data applications. For this benchmark, EEMBC will utilize the popular Memcached program and provide web workloads that mimic real-world scenarios. The benchmark could report metrics such as latency and transactions per second.