ABSTRACT
Recent advances in circuit and process technologies have pushed non-volatile memory technologies into a new era. These technologies exhibit appealing properties such as low power consumption, non-volatility, shock-resistivity, and high density. However, there are challenges to which we need answers in the road of applying non-volatile memories as main memory in computer systems. First, non-volatile memories have limited number of write/erase cycles compared with DRAM memory. Second, write activities on non-volatile memory are more expensive than DRAM memory in terms of energy consumption and access latency. Both challenges will benefit from reduction of the write activities on the nonvolatile memory.
In this paper, we target embedded Chip Multiprocessors (CMPs) with Scratch Pad Memory (SPM) and non-volatile main memory. We introduce data migration and recomputation techniques to reduce the number of write activities on non-volatile memories. Experimental results show that the proposed methods can reduce the number of writes by 59.41% on average, which means that the non-volatile memory can last 2.8 times as long as before. Meanwhile, the finish time of programs is reduced by 31.81% on average.
- N. Eisley, L.-S. Peh, and L. Shang. Leveraging on-chip networks for data cache migration in chip multiprocessors. In PACT '08, pages 197--207, Toronto, Ontario, Canada, 2008. Google ScholarDigital Library
- H. P. Hofstee. Power efficient processor architecture and the cell processor. In HPCA '05, pages 258--262, San Francisco, California, USA, 2005. Google ScholarDigital Library
- M. Kandemir, G. Chen, F. Li, and I. Demirkiran. Using data replication to reduce communication energy on chip multiprocessors. In ASP-DAC '05, pages 769--772, Shanghai, China, 2005. Google ScholarDigital Library
- S. Kaneko and etc. A 600-mhz single-chip multiprocessor with 4.8-gb/s internal shared pipelined bus and 512-kb internal memory. IEEE Journal of Solid-State Circuits, 39(1):184--193, Jan. 2004.Google ScholarCross Ref
- M. Kanellos. Ibm changes directions in magnetic memory, August 2007. http://news.cnet.com/IBM-changes-directions-in-magnetic-memory/2100-1004_3-6203198.Google Scholar
- D.-H. Kang and etc. Two-bit cell operation in diode-switch phase change memory cells with 90nm technology. In Symposium on VLSI Technology, pages 98--99, 2008.Google ScholarCross Ref
- H. Koc, M. Kandemir, E. Ercanli, and O. Ozturk. Reducing off-chip memory access costs using data recomputation in embedded chip multi-processors. In DAC '07, pages 224--229, San Diego, California, 2007. Google ScholarDigital Library
- K. Lee and A. Orailoglu. Application specific non-volatile primary memory for embedded systems. In CODES/ISSS '08, pages 31--36, Atlanta, GA, USA, 2008. Google ScholarDigital Library
- J. Li, P. Ndai, A. Goel, H. Liu, and K. Roy. An alternate design paradigm for robust spin-torque transfer magnetic ram (stt mram) from circuit/architecture perspective. In ASP-DAC '09, pages 841--846, Yokohama, Japan, 2009. Google ScholarDigital Library
- C. Park, J. Seo, S. Bae, H. Kim, S. Kim, and B. Kim. A low-cost memory architecture with nand xip for mobile embedded systems. In CODES+ISSS '03, pages 138--143, Newport Beach, CA, USA, 2003. Google ScholarDigital Library
- D. Roberts, T. Kgil, and T. N. Mudge. Using non-volatile memory to save energy in servers. In DATE '09, pages 743--748, Nice Acropolis, France, 2009. Google ScholarDigital Library
- I. Williams. Phase change memory is another step closer, Oct. 2009. http://www.hpcwire.com/news/Phase-Change-Memory-is-Another-Step-Closer.html.Google Scholar
- C. Q. Xu, C. J. Xue, J. Hu, and E. H.-M. Sha. Optimizing scheduling and intercluster connection for application-specific dsp processors. IEEE TSP, 57(11):4538--4547, 2009. Google ScholarDigital Library
- C. J. Xue, J. Hu, Z. Shao, and E. Sha. Iterational retiming with partitioning: Loop scheduling with complete memory latency hiding. ACM TECS, 9(3):1--26, 2010. Google ScholarDigital Library
- C. J. Xue, Z. Jia, Z. Shao, M. Wang, and E. H.-M. Sha. Optimized address assignment with array and loop transformations for minimizing schedule length. IEEE TCAS, 55(1):379--389, 2008.Google Scholar
- C. J. Xue, Z. Shao, Q. Zhuge, B. Xiao, M. Liu, and E. H.-M. Sha. Optimizing address assignment for scheduling dsps with multiple functional units. IEEE TCAS, 53(9):976--980, 2006.Google Scholar
- F. Yeung and et al. ge2sb2te5 confined structures and integration of 64mb phase-change random access memory. Japanese Journal of Applied Physics, pages 2691--2695, 2005.Google Scholar
- P. Zhou, B. Zhao, J. Yang, and Y. Zhang. A durable and energy efficient main memory using phase change memory technology. In ISCA '09, Austin, Texas, USA, 2009. Google ScholarDigital Library
Index Terms
- Reducing write activities on non-volatile memories in embedded CMPs via data migration and recomputation
Recommendations
Write activity reduction on non-volatile main memories for embedded chip multiprocessors
Recent advances in circuit and semiconductor technologies have pushed Non-Volatile Memory (NVM) technologies into a new era. These technologies exhibit appealing properties such as low power consumption, non-volatility, shock-resistivity, and high ...
Write Activity Minimization for Nonvolatile Main Memory Via Scheduling and Recomputation
Nonvolatile memories such as Flash memory, phase change memory (PCM), and magnetic random access memory (MRAM) have many desirable characteristics for embedded systems to employ them as main memory. However, there are two common challenges we need to ...
Embedded non-volatile memories
SBCCI '07: Proceedings of the 20th annual conference on Integrated circuits and systems designThis tutorial covers trends in embedded non-volatile memories including details of issues for scaling NAND and NOR flash and descriptions of scaled flash memory technologies and various evolutionary flash memory technologies such as trapping site ...
Comments