Control-theoretic techniques and thermal-RC modeling for accurate and localized dynamic thermal management

By Kevin Skadron, Tarek F. Abdelzaher, Mircea R. Stan. Proceedings Eighth International Symposium on High Performance Computer Architecture, 2002

434 citations recorded by the source corpus.

Subfields: Electrical and Electronic Engineering, Hardware and Architecture

View Paper: https://www.semanticscholar.org/paper/0ad82bc5cac34afcee9e7bbe43e9f3e2c786f98f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Electrical and Electronic Engineering

In Hardware and Architecture

Citations in the corpus

Heat-and-run: leveraging SMT and CMP to manage power density through the operating system (2004, methodology): “Other power-density techniques have been applied to supers lar processors, such as dynamic frequency scaling (DFS) a dynamic voltage scaling (DVS) [1, 8, 9].”
Real Time Scheduling Theory: A Historical Perspective (2004, neutral): “Successful control schemes were developed for high-performance servers [71, 79, 151], multimedia streaming [164, 228], real-time databases [12], web storage systems [144], network routers [110, 63], active queue management schemes [98, 99], and processor architecture [204] to name a few.”
Reducing power density through activity migration (2003, methodology): “Junction temperature is modelled by a simple first-order equivalent RC circuit (Figure 1) using the well-known analogy between electrical circuits and thermal models [10].”
Temperature and supply Voltage aware performance and power modeling at microarchitecture level (2005, background): “HotSpot [18], [19] provides a detailed thermal model based on an equivalent distributed circuit of thermal resistances and capacitances that correspond to microarchitectural units and the package with heat spreader and heatsink.”
Temperature-aware microarchitecture: Modeling and implementation (2004, methodology): “Our first work [Skadron et al. 2002] omitted the lateral resistances and the package, both essential for accurate temperatures and time evolution; and used a rather extreme die thickness of 0.”
Techniques for Multicore Thermal Management: Classification and New Exploration (2006, methodology): “In fact, our proportional constant is set two orders of magnitude smaller than that used in [32], in or-1063-6897/06 $20.00 (c) 2006 IEEE der to maintain control with smoother transitions.”
Impact of self-heating effect on long-term reliability and performance degradation in CMOS circuits (2006, background): “Recently performed fast transient thermal analysis of high-performance circuits have shown that different blocks of microprocessors can have the thermal time constants in the range of tens to hundreds microseconds under normal operating conditions [19].”
HotLeakage: A Temperature-Aware Model of Subthreshold and Gate Leakage for Architects (2003, background): “as well as broader processor-design issues like thermal management [4, 17, 25 , 26] manipulate parameters like VT and temperature that are hidden in kdesign or Îleak and this makes it more difficult both to reason about these effects (there is an exponential dependence of leakage on VT , for example”
Temperature-aware microarchitecture (2003, methodology): “Some architectural techniques have already been proposed [4, 13, 16, 22, 26 ], so there is clearly interest in this topic within the architecture field.”
Dynamo: Facebook's Data Center-Wide Power Management System (2016, background): “Based on their control goals, they fall into the following groups: peak power management approaches [4], [8], [26]; thermal management approaches [9], [10], [27]; and average power or energy management approaches [28], [29], [30], [32], [33].”
Server-Level Power Control (2007, background): “Similar techniques for system-level power constraints have been less studied.”