An energy-efficient and error-resilient server ecosystem exceeding conservative scaling limits

Georgios Karakonstantis, Konstantinos Tovletoglou, Lev Mukhanov, Hans Vandierendonck, Dimitrios S. Nikolopoulos, Peter Lawthers, Panos Koutsovasilis, Manolis Maroudas, Christos D. Antonopoulos, Christos Kalogirou, Nikos Bellas, Spyros Lalis, Srikumar Venugopal, Arnau Prat-Perez, Alejandro Lampropoulos, Marios Kleanthous, Andreas Diavastos, Zacharias Hadjilambrou, Panagiota Nikolaou, Yiannakis SazeidesPedro Trancoso, George Papadimitriou, Manolis Kaliorakis, Athanasios Chatzidimitriou, Dimitris Gizopoulos, Shidhartha Das

Research output: Chapter in Book/Report/Conference proceedingConference contribution

15 Citations (Scopus)
442 Downloads (Pure)

Abstract

The explosive growth of Internet-connected devices will soon result in a flood of generated data, which will increase the demand for network bandwidth as well as compute power to process the generated data. Consequently, there is a need for more energy efficient servers to empower traditional centralized Cloud data-centers as well as emerging decentralized data-centers at the Edges of the Cloud. In this paper, we present our approach, which aims at developing a new class of micro-servers - the UniServer - that exceed the conservative energy and performance scaling boundaries by introducing novel mechanisms at all layers of the design stack. The main idea lies on the realization of the intrinsic hardware heterogeneity and the development of mechanisms that will automatically expose the unique varying capabilities of each hardware. Low overhead schemes are employed to monitor and predict the hardware behavior and report it to the system software. The system software including a virtualization and resource management layer is responsible for optimizing the system operation in terms of energy or performance, while guaranteeing non-disruptive operation under the extended operating points. Our characterization results on a 64-bit ARMv8 micro-server in 28nm process reveal large voltage margins in terms of Vmin variation among the 8 cores of the CPU chip, among three different sigma chips, and among different benchmarks with the potential to obtain up-to 38.8% energy savings. Similarly, DRAM characterizations show that refresh rate and voltage can be relaxed by 35x and 5%, respectively, leading to 23.2% power savings on average.

Original languageEnglish
Title of host publicationProceedings of the Design, Automation and Test in Europe Conference and Exhibition, DATE 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1099-1104
Number of pages6
Volume2018-January
ISBN (Electronic)9783981926309, 9783981926316
DOIs
Publication statusPublished - 23 Apr 2018
EventDesign Automation & Test in Europe (DATE) 2018 - International Congress Center, Dresden, Germany
Duration: 19 Mar 201823 Mar 2018
https://www.date-conference.com/

Publication series

NameDesign, Automation and Test in Europe: Proceedings
PublisherIEEE
ISSN (Electronic)1558-1101

Conference

ConferenceDesign Automation & Test in Europe (DATE) 2018
Abbreviated titleDATE
Country/TerritoryGermany
CityDresden
Period19/03/201823/03/2018
Internet address

Fingerprint

Dive into the research topics of 'An energy-efficient and error-resilient server ecosystem exceeding conservative scaling limits'. Together they form a unique fingerprint.

Cite this