Advertisement

Adaptive and Scalable Predictive Page Policies for High Core-Count Server CPUs

  • Tameesh Suri
  • Aneesh Aggarwal
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10172)

Abstract

Increasing datacenter compute requirements has led to tremendous growth in the cadence of CPU cores on chip-multiprocessors. With large number of threads running on a single node, it is critical to achieve high memory bandwidth efficiency on large scale CMPs to support continued growth in the number CPU cores. In this paper, we present several mechanisms that improve the memory efficiency by improving the page hit rate for multi-core processors. In particular, we present memory page-policies that dynamically adapt to the runtime workload characteristics and use thread awareness to reduce contention between different memory address streams from the different threads. Unlike contemporary DRAM page policies such as static or timer-based, the proposed framework profiles the memory stream at runtime and uncovers opportunities to close or keep DRAM pages open, resulting in reduced page-conflicts and improved efficiencies. We implement the proposed policies in a cycle-accurate performance model simulating an 8-core processor. Our results show that the proposed adaptive page policies increase performance of high memory bandwidth workloads in SPECint2006 by up to 3%, and can attain 83% average performance relative to a “perfect” page prediction policy. We further show that the performance improvement from the techniques increases with the number of cores and with making the policies thread-aware in a many-core processor. The implementation cost of our techniques is extremely low, an area overhead of only 69 bits, making them extremely attractive for real-life products.

References

  1. 1.
    Aggarwal, A., Suri, T.: Dynamic Memory Page Policy. US Patent 9,378,127 B2 (2016)Google Scholar
  2. 2.
    Awasthi, M., Nellans, D.W., Balasubramonian, R., Davis, A.: Prediction based dram row-buffer management in the many-core era. In: IEEE International Conference on Parallel Architectures and Compilation Techniques (PACT) (2011)Google Scholar
  3. 3.
    Boughton, K., Gill, R.: Everything you always wanted to know about sdram memory but were afraid to ask (2010). http://www.anandtech.com/show/3851/everything-you-always-wanted-to-know-aboutsdram-memory-but-were-afraid-to-ask
  4. 4.
    Dodd, J.: Adaptive page management. US Patent 7,076,617 (2006)Google Scholar
  5. 5.
    Ghasempour, M., Jaleel, A., Garside, J., Luján, M.: HAPPY: hybrid address-based page policy in DRAMs. In: MEMSYS (2006)Google Scholar
  6. 6.
    Huan, D., Li, Z., Hu, W., Liu, Z.: Processor directed dynamic page policy. In: Jesshope, C., Egan, C. (eds.) ACSAC 2006. LNCS, vol. 4186, pp. 109–122. Springer, Heidelberg (2006). doi: 10.1007/11859802_10 CrossRefGoogle Scholar
  7. 7.
    Jacob, B., Ng, S., Wang, D.: Memory Systems: Cache, DRAM, Disk. Morgan Kaufmann, CA (2010)Google Scholar
  8. 8.
    Kaseridis, D., Stuecheli, J., John, L.K.: Minimalist open-page: a dram page-mode scheduling policy for the many-core era. In: 44th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) (2011)Google Scholar
  9. 9.
    Ma, C., Chen, S.: A dram precharge policy based on address analysis. In: Digital System Design Architectures, Methods and Tools (DSD) (2007)Google Scholar
  10. 10.
    Miura, S., Ayukawa, K., Watanabe, T.: A dynamic-sdram-mode-control scheme for low-power systems with a 32-bit risc cpu. In: International Symposium on Low Power Electronics and Design (ISLPED) (2001)Google Scholar
  11. 11.
    Park, S.I., Park, I.C.: History-based memory mode prediction for improving memory performance. In: International Symposium on Circuits and Systems (2003)Google Scholar
  12. 12.
    Stankovic, V., Milenkovic, N.: Access latency reduction in contemporary dram memories. Facta universitatis (NIS) (2004)Google Scholar
  13. 13.
    Stankovic, V., Milenkovic, N.: Dram controller with a complete predictor. In: 7th International Conference on Telecommunications in Modern Satellite, Cable and Broadcasting Services (2005)Google Scholar
  14. 14.
    Stankovic, V., Milenkovic, N.: Dram controller with a close-page predictor. In: EUROCON (2005)Google Scholar
  15. 15.
    Xu, Y., Agarwal, Aabhas, S., Davis, Brian, T.: Prediction in dynamic SDRAM controller policies. In: Bertels, K., Dimopoulos, N., Silvano, C., Wong, S. (eds.) SAMOS 2009. LNCS, vol. 5657, pp. 128–138. Springer, Heidelberg (2009). doi: 10.1007/978-3-642-03138-0_14 CrossRefGoogle Scholar
  16. 16.

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  1. 1.Intel CorporationSanta ClaraUSA

Personalised recommendations