Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387378083> ?p ?o ?g. }
Showing items 1 to 80 of
80
with 100 items per page.
- W4387378083 abstract "Thermal optimization of a heterogeneous clustered multi-core processor under user-defined quality of service (QoS) targets requires application migration and dynamic voltage and frequency scaling (DVFS). However, selecting the core to execute each application and the voltage/frequency (V/f) levels of each cluster is a complex problem because 1) the diverse characteristics and QoS targets of applications require different optimizations, and 2) per-cluster DVFS requires a global optimization considering all running applications. State-of-the-art resource management for power or temperature minimization either relies on measurements that are commonly not available (such as power) or fails to consider all the dimensions of the optimization (e.g., by using simplified analytical models). To solve this, machine learning (ML) methods can be employed. In particular, imitation learning (IL) leverages the optimality of an oracle policy, yet at low run-time overhead, by training a model from oracle demonstrations. We are the first to employ IL for temperature minimization under QoS targets. We tackle the complexity by training neural network (NN) at design time and accelerate the run-time NN inference using a neural processing unit (NPU). While such NN accelerators are becoming increasingly widespread, they are so far only used to accelerate user applications. In contrast, we use for the first time an existing accelerator on a real platform to accelerate NN-based resource management. To show the superiority of IL compared to reinforcement learning (RL) in our targeted problem, we also develop multi-agent RL-based management. Our evaluation on a HiKey 970 board with an Arm big.LITTLE CPU and NPU shows that IL achieves significant temperature reductions at a negligible run-time overhead. We compare TOP-IL against several techniques. Compared to ondemand Linux governor, TOP-IL reduces the average temperature by up to 17 °C at minimal QoS violations for both techniques. Compared to the RL policy, our TOP-IL achieves 63 % to 89 % fewer QoS violations while resulting similar average temperatures. Moreover, TOP-IL outperforms the RL policy in terms of stability. We additionally show that our IL-based technique also generalizes to different software (unseen applications) and even hardware (different cooling) than used for training." @default.
- W4387378083 created "2023-10-06" @default.
- W4387378083 creator A5017470419 @default.
- W4387378083 creator A5035265423 @default.
- W4387378083 creator A5037770816 @default.
- W4387378083 creator A5063508488 @default.
- W4387378083 date "2023-10-05" @default.
- W4387378083 modified "2023-10-16" @default.
- W4387378083 title "NPU-Accelerated Imitation Learningfor Thermal Optimizationof QoS-Constrained Heterogeneous Multi-Cores" @default.
- W4387378083 cites W1974386611 @default.
- W4387378083 cites W2003756933 @default.
- W4387378083 cites W2475534957 @default.
- W4387378083 cites W2768195099 @default.
- W4387378083 cites W2770465907 @default.
- W4387378083 cites W2889579855 @default.
- W4387378083 cites W2896609748 @default.
- W4387378083 cites W2968713519 @default.
- W4387378083 cites W2980002279 @default.
- W4387378083 cites W3013222071 @default.
- W4387378083 cites W3014326829 @default.
- W4387378083 cites W3021369463 @default.
- W4387378083 cites W3195447397 @default.
- W4387378083 doi "https://doi.org/10.1145/3626320" @default.
- W4387378083 hasPublicationYear "2023" @default.
- W4387378083 type Work @default.
- W4387378083 citedByCount "0" @default.
- W4387378083 crossrefType "journal-article" @default.
- W4387378083 hasAuthorship W4387378083A5017470419 @default.
- W4387378083 hasAuthorship W4387378083A5035265423 @default.
- W4387378083 hasAuthorship W4387378083A5037770816 @default.
- W4387378083 hasAuthorship W4387378083A5063508488 @default.
- W4387378083 hasBestOaLocation W43873780831 @default.
- W4387378083 hasConcept C111919701 @default.
- W4387378083 hasConcept C115903868 @default.
- W4387378083 hasConcept C120314980 @default.
- W4387378083 hasConcept C121332964 @default.
- W4387378083 hasConcept C154945302 @default.
- W4387378083 hasConcept C157742956 @default.
- W4387378083 hasConcept C163258240 @default.
- W4387378083 hasConcept C2776214188 @default.
- W4387378083 hasConcept C2779960059 @default.
- W4387378083 hasConcept C31258907 @default.
- W4387378083 hasConcept C41008148 @default.
- W4387378083 hasConcept C50644808 @default.
- W4387378083 hasConcept C5119721 @default.
- W4387378083 hasConcept C55166926 @default.
- W4387378083 hasConcept C62520636 @default.
- W4387378083 hasConcept C97541855 @default.
- W4387378083 hasConceptScore W4387378083C111919701 @default.
- W4387378083 hasConceptScore W4387378083C115903868 @default.
- W4387378083 hasConceptScore W4387378083C120314980 @default.
- W4387378083 hasConceptScore W4387378083C121332964 @default.
- W4387378083 hasConceptScore W4387378083C154945302 @default.
- W4387378083 hasConceptScore W4387378083C157742956 @default.
- W4387378083 hasConceptScore W4387378083C163258240 @default.
- W4387378083 hasConceptScore W4387378083C2776214188 @default.
- W4387378083 hasConceptScore W4387378083C2779960059 @default.
- W4387378083 hasConceptScore W4387378083C31258907 @default.
- W4387378083 hasConceptScore W4387378083C41008148 @default.
- W4387378083 hasConceptScore W4387378083C50644808 @default.
- W4387378083 hasConceptScore W4387378083C5119721 @default.
- W4387378083 hasConceptScore W4387378083C55166926 @default.
- W4387378083 hasConceptScore W4387378083C62520636 @default.
- W4387378083 hasConceptScore W4387378083C97541855 @default.
- W4387378083 hasLocation W43873780831 @default.
- W4387378083 hasOpenAccess W4387378083 @default.
- W4387378083 hasPrimaryLocation W43873780831 @default.
- W4387378083 hasRelatedWork W1967088250 @default.
- W4387378083 hasRelatedWork W2045932760 @default.
- W4387378083 hasRelatedWork W2073713056 @default.
- W4387378083 hasRelatedWork W2110441383 @default.
- W4387378083 hasRelatedWork W2125620709 @default.
- W4387378083 hasRelatedWork W2914522629 @default.
- W4387378083 hasRelatedWork W3110702597 @default.
- W4387378083 hasRelatedWork W4306904969 @default.
- W4387378083 hasRelatedWork W4362501864 @default.
- W4387378083 hasRelatedWork W4380318855 @default.
- W4387378083 isParatext "false" @default.
- W4387378083 isRetracted "false" @default.
- W4387378083 workType "article" @default.