Abstract
Recent research on mitigating thermal problems in 3D memories has covered reactive strategies that reduce memory power consumption, and thereby, performance, when the memory temperature reaches the maximum operating limit. Such techniques could benefit from temperature prediction and avoid unnecessary invocations and state transitions of the thermal management strategy. We develop an accurate steady state temperature predictor for thermal management of 3D memories. We utilize the symmetries in the floorplan, along with other design insights, to reduce the predictor’s model parameters, making it lightweight and suitable for runtime thermal management. Using the temperature prediction, we introduce PredictNcool, a proactive thermal management strategy to reduce application runtime and memory energy. We compare PredictNcool with two recent thermal management strategies and our experiments show that the proposed optimization results in performance improvements of 28% and 5%, and memory subsystem energy reductions of 38% and 12% (on average).
- T. Adegbija and A. Gordon-Ross. 2018. TaPT: Temperature-aware dynamic cache optimization for embedded systems. Computers (2018).Google Scholar
- R. Ayoub, K. R. Indukuri, and T. S. Rosing. 2010. Energy efficient proactive thermal management in memory subsystem. In ISLPED.Google Scholar
- R. Ayoub and A. Orailoglu. 2010. Performance and energy efficient cache migration approach for thermal management in embedded systems. In GLSVLSI.Google Scholar
- P. Bogdan, P. P. Pande, H. Amrouch, M. Shafique, and J. Henkel. 2016. Power and thermal management in massive multicore chips: Theoretical foundation meets architectural innovation and resource allocation. In CASES.Google Scholar
- D. Brooks and M. Martonosi. 2001. Dynamic thermal management for high-performance microprocessors. In HPCA.Google Scholar
- D. Calvo, P. González, L. Díaz, H. Posadas, P. Sánchez, E. Villar, A. Acquaviva, and E. Macii. 2011. A multi-processing systems-on-chip native simulation framework for power and thermal-aware design. JOLPE (2011).Google Scholar
- T. E. Carlson, W. Heirman, S. Eyerman, I. Hur, and L. Eeckhout. 2014. An evaluation of high-level mechanistic core models. TACO (2014).Google Scholar
- K. Chen et al. 2012. CACTI-3DD: Architecture-level modeling for 3D die-stacked DRAM main memory. In DATE.Google Scholar
- R. Cochran and S. Reda. 2013. Thermal prediction and adaptive control through workload phase detection. TODAES (2013).Google Scholar
- A. K. Coskun, T. S. Rosing, and K. C Gross. 2008. Proactive temperature management in MPSoCs. In ISLPED.Google Scholar
- D. Cuesta, J. Ayala, J. Hidalgo, M. Poncino, A. Acquaviva, and E. Macii. 2010. Thermal-aware floorplanning exploration for 3D multi-core architectures. In GLSVLSI.Google Scholar
- K. Dev, A. N. Nowroz, and S. Reda. 2013. Power mapping and modeling of multi-core processors. In ISLPED.Google Scholar
- A. Fourmigue, G. Beltrame, and G. Nicolescu. 2014. Efficient transient thermal simulation of 3D ICs with liquid-cooling and through silicon vias. In DATE.Google Scholar
- M. H. Hajkazemi et al. 2017. Heterogeneous HMC+DDRx memory management for performance-temperature tradeoffs. JETCS (Sept. 2017).Google Scholar
- F. Hameed, M. A. A. Faruque, and J. Henkel. 2011. Dynamic thermal management in 3D multi-core architecture through run-time adaptation. In DATE.Google Scholar
- J. L. Henning. 2006. SPEC CPU2006 benchmark descriptions. ACM SIGARCH Computer Architecture News (2006).Google Scholar
- H. Homayoun, A. Gupta, A. Veidenbaum, A. Sasan, F. Kurdahi, and N. Dutt. 2010. RELOCATE: Register file local access pattern redistribution mechanism for power and thermal management in out-of-order embedded processor. In HiPEAC.Google Scholar
- H. Huang, K. G. Shin, C. Lefurgy, and T. Keller. 2005. Improving energy efficiency by making DRAM less randomly accessed. In ISLPED.Google Scholar
- J. Jeddeloh and B. Keeth. 2012. Hybrid memory cube new DRAM architecture increases density and performance. In VLSIT.Google Scholar
- D. Juan et al. 2014. Statistical peak temperature prediction and thermal yield improvement for 3D chip multiprocessors. TODAES (2014).Google Scholar
- D. Juan, H. Zhou, D. Marculescu, and X. Li. 2012. A learning-based autoregressive model for fast transient thermal analysis of chip-multiprocessors. In ASPDAC.Google Scholar
- P. Kumar and D. Atienza. 2010. Neural network based on-chip thermal simulator. In ISCAS.Google Scholar
- D. Lee, S. Das, J. R. Doppa, P. P. Pande, and K. Chakrabarty. 2018. Performance and thermal tradeoffs for energy-efficient monolithic 3D network-on-chip. TODAES (2018).Google Scholar
- C. H. Liao, C. H. P. Wen, and K. Chakrabarty. 2015. An online thermal-constrained task scheduler for 3D multi-core processors. In DATE’15.Google Scholar
- W. Liu, L. Yang, W. Jiang, L. Feng, N. Guan, W. Zhang, and N. Dutt. 2018. Thermal-aware task mapping on dynamically reconfigurable network-on-chip based multiprocessor system-on-chip. TC (2018).Google Scholar
- W. Lo et al. 2016. Thermal-aware dynamic page allocation policy by future access patterns for Hybrid Memory Cube (HMC). In DATE.Google Scholar
- G. L. Loi et al. 2006. A thermally-aware performance analysis of vertically integrated (3D) processor-memory hierarchy. In DAC.Google Scholar
- Y. Lu et al. 2016. Rank-aware dynamic migrations and adaptive demotions for DRAM power management. TC (2016).Google Scholar
- J. Meng et al. 2012. Optimizing energy efficiency of 3-D multicore systems with stacked DRAM under power and thermal constraints. In DAC.Google Scholar
- F. Pedregosa et al. 2011. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research 12 (2011), 2825--2830.Google Scholar
Digital Library
- L. Siddhu and P. R. Panda. 2019. FastCool: Leakage aware dynamic thermal management of 3D memories. In DATE.Google Scholar
- G. Singla, G. Kaur, A. K. Unver, and U. Y. Ogras. 2015. Predictive dynamic thermal and power management for heterogeneous mobile platforms. In DATE.Google Scholar
- F. Sironi, M. Maggio, R. Cattaneo, G. F. D. Nero, D. Sciuto, and M. D. Santambrogio. 2013. ThermOS: System support for dynamic thermal management of chip multi-processors. In PACT.Google Scholar
- I. G. Thakkar, S. Pasricha, et al. 2018. LIBRA: Thermal and process variation aware reliability management in photonic networks-on-chip. TMSCS (2018).Google Scholar
- S. Xydis, G. Palermo, and C. Silvano. 2013. Thermal-aware datapath merging for coarse-grained reconfigurable processors. In DATE.Google Scholar
- Y. Ye, Z. Wang, P. Yang, J. Xu, X. Wu, X. Wang, M. Nikdast, Z. Wang, and L. H. K. Duong. 2014. System-level modeling and analysis of thermal effects in WDM-based optical networks-on-chip. TCAD (2014).Google Scholar
- M. Zapater et al. 2013. Leakage and temperature aware server control for improving energy efficiency in data centers. In DATE.Google Scholar
- K. Zhang, A. Guliani, S. Ogrenci-Memik, G. Memik, K. Yoshii, R. Sankaran, and P. Beckman. 2018. Machine learning-based temperature prediction for runtime thermal management across system components. TPDS (2018).Google Scholar
- R. Zhang, M. R. Stan, and K. Skadron. 2015. HotSpot 6.0: Validation, acceleration and extension. (2015).Google Scholar
- Z. Zhao, A. Gerstlauer, and L. K. John. 2017. Source-level performance, energy, reliability, power and thermal (PERPT) simulation. TCAD (2017).Google Scholar
- J. Zheng, N. Wu, L. Zhou, Y. Ye, and K. Sun. 2016. DFSB-based thermal management scheme for 3D NoC-bus architectures. TVLSI (2016).Google Scholar
Index Terms
PredictNcool: Leakage Aware Thermal Management for 3D Memories Using a Lightweight Temperature Predictor
Recommendations
CoMeT: An Integrated Interval Thermal Simulation Toolchain for 2D, 2.5D, and 3D Processor-Memory Systems
Processing cores and the accompanying main memory working in tandem enable modern processors. Dissipating heat produced from computation remains a significant problem for processors. Therefore, the thermal management of processors continues to be an ...
Leakage-Aware Dynamic Thermal Management of 3D Memories
3D memory systems offer several advantages in terms of area, bandwidth, and energy efficiency. However, thermal issues arising out of higher power densities have limited their widespread use. While prior works have looked at reducing dynamic power ...
Opportunities of synergistically adjusting voltage-frequency levels of cores and DRAMs in CMPs with 3d-stacked DRAMs for efficient thermal control
Chip-Multiprocessors (CMPs) with 3D-stacked DRAMs is promising for solving the memory wall problem, but the high power density makes 3D ICs frequently operate at or near the thermal limit. System hot spot of a CMP with 3Dstacked DRAMs is usually in ...






Comments