Predictive dynamic thermal management for multicore systems

Abstract

Recently, processor power density has been increasing at an alarming rate resulting in high on-chip temperature. Higher temperature increases current leakage and causes poor reliability. In this paper, we propose a Predictive Dynamic Thermal Management (PDTM) based on Application-based Thermal Model (ABTM) and Core-based Thermal Model (CBTM) in the multicore systems. ABTM predicts future temperature based on the application specific thermal behavior, while CBTM estimates core temperature pattern by steady state temperature and workload. The accuracy of our prediction model is 1.6% error in average compared to the model in HybDTM [8], which has at most 5% error. Based on predicted temperature from ABTM and CBTM, the proposed PDTM can maintain the system temperature below a desired level by moving the running application from the possible overheated core to the future coolest core (migration) and reducing the processor resources (priority scheduling) within multicore systems. PDTM enables the exploration of the tradeoff between throughput and fairness in temperature-constrained multicore systems. We implement PDTM on Intel's Quad-Core system with a specific device driver to access Digital Thermal Sensor (DTS). Compared against Linux standard scheduler, PDTM can decrease average temperature about 10%, and peak temperature by 5• C with negligible impact of performance under 1%, while running single SPEC2006 benchmark. Moreover, our PDTM outperforms HRTM [10] in reducing average temperature by about 7% and peak temperature by about 3• C with performance overhead by 0.15% when running single benchmark.

Compact thermal modeling for packaged microprocessor design with practical power maps

“…To mitigate this problem, we use the average temperature of any instance time to guide model switching. The thermal gradients in a well-designed chip are typically well managed and reduced by the online thermal management techniques [33,20]. Even with some degrees of thermal gradients, the local models should still be valid as it is a localized model and should be valid for a temperature range as long as the average temperature is still representative for the overall temperature of the processor die.…”

Thermal/performance trade-off in network-on-chip architectures

“…Indeed, several authors provide a methodology based on simple temperature predictive control to avoid exceeding a predefined threshold value [28]. History-based approaches in this sense have been proposed in [30]. The performance impact of many DTM techniques for high-performance microprocessors has been extensively discussed in [6].…”

Physical-aware predictive dynamic thermal management of multi-core processors

“…In response to these challenges, various Dynamic Thermal Management (DTM) techniques have been proposed to mitigate the thermal concerns of processors [2][3][4][5][6][7][8][9][10][15][16][17][18][20][21][22][23]. DTM is a set of techniques that control processor temperature at run-time so that temperature does not go beyond the critical temperature threshold.…”

“…Moreover, reading different performance counters imposes significant overhead on application execution at run-time [1]. Therefore, recently proposed methods model overall core and application temperature with the aid of physical sensor and steady state temperature [16][17].…”

“…In order to have more efficient DTM techniques, recent works predict future temperature of cores to reduce overheat temperature with negligible performance overhead [16][17]. Their proactive task migration approaches predict the future temperature and manage the workload to reduce and balance the temperature before reaching the temperature threshold.…”