The Community for Technology Leaders
Green Image
Issue No. 03 - May-June (2016 vol. 36)
ISSN: 0272-1732
pp: 94-104
Sheng Li , Intel Labs
Hyeontaek Lim , Carnegie Mellon University
Victor W. Lee , Intel Labs
Jung Ho Ahn , Seoul National University
Anuj Kalia , Carnegie Mellon University
Michael Kaminsky , Intel Labs
David G. Andersen , Carnegie Mellon University
Seongil O , Seoul National University
Sukhan Lee , Seoul National University
Pradeep Dubey , Intel Labs
Distributed in-memory key-value stores (KVSs) have become a critical data-serving layer in cloud computing and big data infrastructure. Unfortunately, KVSs have demonstrated a gap between achieved and available performance, QoS, and energy efficiency on commodity platforms. Two research thrusts have focused on improving key-value performance: hardware-centric research has started to explore specialized platforms for KVSs, and software-centric research revisited the KVS application to address fundamental software bottlenecks. Unlike prior research focusing on hardware or software in isolation, the authors aimed to full-stack (software through hardware) architect high-performance and efficient KVS platforms. Their full-system characterization identifies the critical hardware/software ingredients for high-performance KVS systems and suggests optimizations to achieve record-setting performance and energy efficiency: 120~167 million requests per second (RPS) on a single commodity server. They propose a future many-core platform and via detailed simulations demonstrate the capability of achieving a billion RPS with a single server platform.
Servers, Key value systems, Concurrency control, Memory management, Program processors, Performance evaluation, Field programmable analog arrays
Sheng Li, Hyeontaek Lim, Victor W. Lee, Jung Ho Ahn, Anuj Kalia, Michael Kaminsky, David G. Andersen, Seongil O, Sukhan Lee, Pradeep Dubey, "Achieving One Billion Key-Value Requests per Second on a Single Server", IEEE Micro, vol. 36, no. , pp. 94-104, May-June 2016, doi:10.1109/MM.2016.13
100 ms
(Ver 3.3 (11022016))