On-Line Global Energy Optimization in Multi-Core Systems Using Principles of Analog Computation

This work presents the design and the silicon implementation of an on-line energy optimizer unit, which is capable of dynamically adjusting power supply voltages and operating frequencies of multiple processing elements. The optimized voltage/frequency assignments are tailored to the instantaneous workload information and fully adaptive to variations in process and temperature. The optimizer unit has a fast response time of 50 ?s, occupies a silicon area of 0.021mm2 / task and dissipates 2 mW / task.

[1]  Eric A. Vittoz,et al.  Analog VLSI signal processing: Why, where, and how? , 1994, J. VLSI Signal Process..

[2]  B. Gilbert.  Translinear circuits: a proposed classification , 1975 .

[3]  D.M. Monticelli.  Taking a system approach to energy management , 2003, ESSCIRC 2004 - 29th European Solid-State Circuits Conference (IEEE Cat. No.03EX705).

[4]  Luca Benini,et al.  Event-driven power management , 2001, IEEE Trans. Comput. Aided Des. Integr. Circuits Syst..

[5]  Alper Demir,et al.  Characterizing and exploiting task load variability and correlation for energy management in multi core systems , 2005, 3rd Workshop on Embedded Systems for Real-Time Multimedia, 2005..

[6]  Niraj K. Jha,et al.  Joint dynamic voltage scaling and adaptive body biasing for heterogeneous distributed real-time embedded systems , 2003, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[7]  Xiaobo Sharon Hu,et al.  Task scheduling and voltage selection for energy minimization , 2002, DAC '02.

[8]  Trevor Mudge,et al.  Combined dynamic voltage scaling and adaptive body biasing for lower power microprocessors under dynamic workloads , 2002, ICCAD 2002.

[9]  Rajesh K. Gupta,et al.  Dynamic voltage scaling for systemwide energy minimization in real-time embedded systems , 2004, Proceedings of the 2004 International Symposium on Low Power Electronics and Design (IEEE Cat. No.04TH8758).

[10]  Eric A. Vittoz,et al.  Linear networks based on transistors , 1993 .

[11]  Krisztian Flautner,et al.  A Combined Hardware-Software Approach for Low-Power SoCs: Applying Adaptive Voltage Scaling and Intelligent Energy Management Software , 2003 .

[12]  Petru Eles,et al.  Overhead-conscious voltage selection for dynamic and leakage energy reduction of time-constrained systems , 2004, Proceedings Design, Automation and Test in Europe Conference and Exhibition.

[13]  Eric A. Vittoz.  Pseudo-Resistive Networks and their Applications to Analog Collective Computation , 1997, ICANN.

[14]  Luca Benini,et al.  A survey of design techniques for system-level dynamic power management , 2000, IEEE Trans. Very Large Scale Integr. Syst..

[15]  Zeynep Toprak Deniz.  Multi-unit global energy management and optimization for network-on-chip applications , 2006 .

[16]  E. A. Vittoz.  Analog VLSI for collective computation , 1998, 1998 IEEE International Conference on Electronics, Circuits and Systems. Surfing the Waves of Science and Technology (Cat. No.98EX196).

[17]  J. Maxwell.  A Treatise on Electricity and Magnetism , 1873, Nature.

[18]  Xiaobo Sharon Hu,et al.  Energy minimization of real-time tasks on variable voltage processors with transition energy overhead , 2003, ASP-DAC '03.

[19]  Christian Enz,et al.  A Basic Property of MOS Transistors and its Circuit Implications , 2005 .