This Article 
 Bibliographic References 
 Add to: 
Cache Coherence in Intelligent Memory Systems
July 2003 (vol. 52 no. 7)
pp. 960-966

Abstract—The Active Pages model of intelligent memory can speed up data-intensive applications by up to two to three orders of magnitude over conventional systems. A fundamental problem with intelligent memory, however, arises when data cached by the processor is modified by logic in the memory. The Active Page model inherently limits sharing, keeping coherence tractable, but exacerbates saturation problems. We first present a hybrid snoopy/directory protocol for use in Active Pages. Limited sharing allows for a low-latency, low-bandwidth hybrid protocol. A transparent remapping mechanism is added for efficient caching. On smaller data sizes, explicit flushing and hardware coherence exhibit similar performance, but hardware coherence is easier to program and uses less bandwidth. Finally, we examine SMP multiprocessor systems to mitigate saturation effects. As the number of threads increases, the bandwidth needs increase, making hardware coherence even more attractive.

[1] D. Burger and T.M. Austin, "The SimpleScalar Toolset, Version 2.0," Computer Architecture News, Vol. 25 No. 3, Jun. 1997, pp. 13-25.
[2] M. Gokhale, B. Holmes, and K. Iobst, "The Terasys Massively Parallel PIM Array," Computer, vol. 28, no. 4, Apr. 1995, pp. 23-31.
[3] K. Itoh, Y. Nakagome, S. Kimura, and T. Watanabe, Limitations and Challenges of Multigigabit DRAM Chip Design IEEE J. Solid-State Circuits, vol. 32, no. 5, pp. 624-634, 1997.
[4] Intel, Intel Architecture Software Developer's Manual, 1997.
[5] D. Kim, M. Chaudhuri, and M. Heinrich, Leveraging Cache Coherence in Active Memory Systems Proc. 16th Int'l Conf. Supercomputing, June 2002.
[6] D.M. Keen, Novel Designs and Uses of Communication in Auxiliary Processing Systems PhD thesis, Univ. of California Davis, 2002.
[7] K. Murakami, S. Shirakawa, and H. Miyajima, "Parallel Processing RAM Chip with 256Mb DRAM and Quad Processor," Dig. Technical Papers, 1997 IEEE Int'l Solid-State Circuits Conf., IEEE, 1997, pp. 228-229.
[8] M. Oskin, F. Chong, and T. Sherwood, “Active Pages: A Computation Model for Intelligent Memory,” Proc. 25th Ann. Int'l Symp. Computer Architecture, pp. 192-203, June 1998.
[9] M. Oskin, J. Hensley, D. Keen, F.T. Chong, M. Farrens, and A. Chopra, Exploiting ILP in Page-Based Intelligent Memory Proc. 32nd Ann. Int'l Symp. Microarchitecture, Nov. 1999.
[10] D. Patterson, T. Anderson, N. Cardwell, R. Fromm, K. Keeton, C. Kozyrakis, R. Thomas, and K. Yelick, A Case for Intelligent DRAM: IRAM IEEE Micro, Apr. 1997.

Index Terms:
Intelligent memory, merged DRAM logic, cache coherence.
Diana Keen, Mark Oskin, Justin Hensley, Frederic T. Chong, "Cache Coherence in Intelligent Memory Systems," IEEE Transactions on Computers, vol. 52, no. 7, pp. 960-966, July 2003, doi:10.1109/TC.2003.1214343
Usage of this product signifies your acceptance of the Terms of Use.