Coordinated Management of Processor Configuration and Cache Partitioning to Optimize Energy under QoS Constraints

Nejat, Mehrzad; Manivannan, Madhavan; Pericàs, Miquel; Stenström, Per

doi:10.1109/ipdps47924.2020.00067

Cited by 4 publications

(7 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(4) It restarts the instruction fetch. A similar mechanism has been previously studied in [31]. In this work, we consider two different core sizes, as detailed in Section 4.2.…”

Section: Resource Control Knobsmentioning

confidence: 94%

“…Core Scaling: Adaptive core architectures have been studied in previous work [2,6,23,24,31] as an effective approach in reducing processor energy consumption. While there are many different adaptive architectures proposed, we consider a simple mechanism with minimal overhead.…”

Section: Resource Control Knobsmentioning

confidence: 99%

“…This process requires N − 1 reductions for an N core system. A similar optimization approach has been previously used in [14,[30][31][32].…”

Section: Global Optimizationmentioning

confidence: 99%

“…We used the Linux perf tool for this measurement. In addition to CSM, we have implemented the algorithm proposed in the previous work [31], which we refer to as C3P. We also extended C3P with a simple local slack management (LSM) algorithm for each core.…”

Section: Algorithmmentioning

confidence: 99%

“…• C3P: We use the RM presented in the state-of-the-art [31] for comparison. We refer to this scheme as "Coordinated Core Configuration and Cache Partitioning" (C3P).…”

Section: Evaluated Rm Algorithmsmentioning

confidence: 99%

See 4 more Smart Citations

Cooperative Slack Management: Saving Energy of Multicore Processors by Trading Performance Slack Between QoS-Constrained Applications

Nejat

Manivannan

Pericàs

et al. 2022

ACM Trans. Archit. Code Optim.

Self Cite

View full text Add to dashboard Cite

Processor resources can be adapted at runtime according to the dynamic behavior of applications to reduce the energy consumption of multicore processors without affecting the Quality-of-Service (QoS). To achieve this, an online resource management scheme is needed to control processor configurations such as cache partitioning, dynamic voltage-frequency scaling, and dynamic adaptation of core resources. Prior State-of-the-art has shown the potential for reducing energy without any performance degradation by coordinating the control of different resources. However, in this article, we show that by allowing short-term variations in processing speed (e.g., instructions per second rate), in a controlled fashion, we can enable substantial improvements in energy savings while maintaining QoS. We keep track of such variations in the form of performance slack. Slack can be generated, at some energy cost, by processing faster than the performance target. On the other hand, it can be utilized to save energy by allowing a temporary relaxation in the performance target. Based on this insight, we present Cooperative Slack Management (CSM). During runtime, CSM finds opportunities to generate slack at low energy cost by estimating the performance and energy for different resource configurations using analytical models. This slack is used later when it enables larger energy savings. CSM performs such trade-offs across multiple applications, which means that the slack collected for one application can be used to reduce the energy consumption of another. This cooperative approach significantly increases the opportunities to reduce system energy compared with independent slack management for each application. For example, we show that CSM can potentially save up to 41% of system energy (on average, 25%) in a scenario in which both prior art and an extended version with local slack management for each core are ineffective.

show abstract

“…(4) It restarts the instruction fetch. A similar mechanism has been previously studied in [31]. In this work, we consider two different core sizes, as detailed in Section 4.2.…”

Section: Resource Control Knobsmentioning

confidence: 94%

Section: Resource Control Knobsmentioning

confidence: 99%

“…This process requires N − 1 reductions for an N core system. A similar optimization approach has been previously used in [14,[30][31][32].…”

Section: Global Optimizationmentioning

confidence: 99%

Section: Algorithmmentioning

confidence: 99%

“…• C3P: We use the RM presented in the state-of-the-art [31] for comparison. We refer to this scheme as "Coordinated Core Configuration and Cache Partitioning" (C3P).…”

Section: Evaluated Rm Algorithmsmentioning

confidence: 99%

See 3 more Smart Citations