Energy Conservation in Memory Hierarchies using Power-Aware Cached-DRAM

Main memory has become one of the largest contributors to overall energy consumption and offers many opportunities for power/energy reduction. In this paper, we propose a new memory organization, called {em Power-Aware Cached-DRAM} (PA-CDRAM), that integrates a moderately sized cache directly into a memory device. We use this cache to turn a memory bank off immediately after a memory access to reduce energy consumption. While other work has used CDRAM to improve memory performance, we modify CDRAM to reduce energy consumption. In this paper, we describe our memory organization and describe the challenges for achieving low energy consumption and how to address them. We evaluate the approach using a cycle accurate processor and memory simulator. Our results show that PA-CDRAM achieves an average 28% improvement in the energy-delay product when compared to a time-out power management technique.

[1]  Wei-Fen Lin,et al.  Reducing DRAM latencies with an integrated memory hierarchy design , 2001, Proceedings HPCA Seventh International Symposium on High-Performance Computer Architecture.

[2]  Mahmut T. Kandemir,et al.  Leakage Current: Moore's Law Meets Static Power , 2003, Computer.

[3]  Mahmut T. Kandemir,et al.  DRAM energy management using software and hardware directed power mode control , 2001, Proceedings HPCA Seventh International Symposium on High-Performance Computer Architecture.

[4]  Kang G. Shin,et al.  Design and Implementation of Power-Aware Virtual Memory , 2003, USENIX ATC, General Track.

[5]  Matthias Gries,et al.  Performance evaluation of recent DRAM architectures for embedded systems , 1999 .

[6]  Massoud Pedram,et al.  Transition reduction in memory buses using sector-based encoding techniques , 2004, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[7]  James E. Smith,et al.  Performance Of Cached Dram Organizations In Vector Supercomputers , 1993, Proceedings of the 20th Annual International Symposium on Computer Architecture.

[8]  Trevor N. Mudge,et al.  High-Performance DRAMs in Workstation Environments , 2001, IEEE Trans. Computers.

[9]  Tajana Simunic,et al.  Energy estimation of peripheral devices in embedded systems , 2004, GLSVLSI '04.

[10]  Duncan G. Elliott,et al.  Computational Ram: A Memory-simd Hybrid And Its Application To Dsp , 1992, 1992 Proceedings of the IEEE Custom Integrated Circuits Conference.

[11]  Mahmut T. Kandemir,et al.  VL-CDRAM: variable line sized cached DRAMs , 2003, First IEEE/ACM/IFIP International Conference on Hardware/ Software Codesign and Systems Synthesis (IEEE Cat. No.03TH8721).

[12]  Norbert Wehn,et al.  Embedded DRAM Development: Technology, Physical Design, and Application Issues , 2001, IEEE Des. Test Comput..

[13]  Alvin R. Lebeck,et al.  Power aware page allocation , 2000, SIGP.

[14]  Mahmut T. Kandemir,et al.  vEC: virtual energy counters , 2001, PASTE '01.

[15]  Gershon Kedem,et al.  WCDRAM: A fully associative integrated Cached-DRAM with wide cache lines , 1997 .

[16]  Mahmut T. Kandemir,et al.  Automatic data migration for reducing energy consumption in multi-bank memory systems , 2002, DAC '02.

[17]  Norman P. Jouppi,et al.  Cacti 3. 0: an integrated cache timing, power, and area model , 2001 .

[18]  Narayanan Vijaykrishnan,et al.  Multiple access caches: Energy implications , 2000, Proceedings IEEE Computer Society Workshop on VLSI 2000. System Design for a System-on-Chip Era.

[19]  Zhao Zhang,et al.  Fine-grain priority scheduling on multi-channel memory systems , 2002, Proceedings Eighth International Symposium on High Performance Computer Architecture.