Leveraging Approximation to Improve Datacenter Resource Efficiency

Cloud multi-tenancy is typically constrained to a single interactive service colocated with one or more batch, low-priority services, whose performance can be sacrificed. Approximate computing applications offer the opportunity to enable tighter colocation among multiple applications whose performan...

Full description

Saved in:
Bibliographic Details
Published inIEEE computer architecture letters Vol. 17; no. 2; pp. 171 - 174
Main Authors Kulkarni, Neeraj, Qi, Feng, Delimitrou, Christina
Format Journal Article
LanguageEnglish
Published New York IEEE 01.07.2018
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Approximate computing
Approximation
Cloud computing
Computation
Interference
Mathematical analysis
Monitoring
Quality of service
Runtime
scheduling and task partitioning
Super (very large) computers
support for dynamic compilation
Switches
Online AccessGet full text

Cover

More Information
Summary:Cloud multi-tenancy is typically constrained to a single interactive service colocated with one or more batch, low-priority services, whose performance can be sacrificed. Approximate computing applications offer the opportunity to enable tighter colocation among multiple applications whose performance is important. We present Pliant, a lightweight cloud runtime that leverages the ability of approximate computing applications to tolerate some loss in output quality to boost the utilization of shared servers. During periods of high contention, Pliant employs incremental and interference-aware approximation to reduce interference in shared resources. We evaluate Pliant across different approximate applications, and show that it preserves QoS for all co-scheduled workloads, while incurring at most a 5 percent loss in output quality.
ISSN:1556-6056
1556-6064
DOI:10.1109/LCA.2018.2845841