Jichuan Chang
Parthasarathy Ranganathan
Trevor Mudge
David Roberts
ABSTRACT
The adoption of non-volatile memories (NVMs) in system architecture and the growth in data-centric workloads offer exciting opportunities for new designs. In this paper, we examine the potential and limit of designs that move compute in close proximity to NVM-based data stores. To address the challenges in evaluating such system architectures for distributed systems, we develop and validate a new methodology for large-scale data-centric workloads. We then study "nanostores" as an example design that constructs distributed systems from building blocks with 3D-stacked compute and NVM layers on the same chip, replacing both traditional storage and memory with NVM. Our limits study demonstrates significant potential of this approach (3-162X improvement in energy delay product) over 2015 baselines, particularly for IO-intensive workloads. We also discuss and quantify the impact of network bandwidth, software scalability, and power density, and design tradeoffs for future NVM-based data-centric architectures.
- M. Mayer. The physics of data. Talk at Xerox PARC, 2009.Google Scholar
- P. Ranganathan, From Microprocessors to Nanostores: Rethinking Data-Centric Systems. IEEE Computer Vol. 44(1), 2011, pp. 39--48. Google ScholarDigital Library
- COTSon: Infrastructure for system-level simulation. MICRO Tutorial, 2008.Google Scholar
- ITRS roadmap. http://www.itrs.net/, 2009.Google Scholar
- D. Andersen, et al. FAWN: A fast array of wimpy nodes. SOSP, 2009. Google ScholarDigital Library
- A. Caulfield, et a;. Gordon: an improved architecture for data-intensive applications. IEEE Micro, 30(1), 2010. Google ScholarDigital Library
- H. Volos, A. Tack, et al. Mnemosyne: Lightweight Persistent Memory. ASPLOS, 2011. Google ScholarDigital Library
- A. Cockcroft. Millicomputing: The future in your pocket and your datacenter. USENIX invited talk, 2008.Google Scholar
- J. Dean and S. Ghemawat. MapReduce: Simplified data processing on large clusters. OSDI, 2004. Google ScholarDigital Library
- J. Ousterhout et. al. The case for RAMCloud. Communications of the ACM, 54(7):121--130, 2011. Google ScholarDigital Library
- T. Kgil et al. PicoServer: Using 3D Stacking Technology To Enable A Compact Energy Efficient Chip Multiprocessor. ASPLOS, 2006. Google ScholarDigital Library
- M. Gokhale, B. Holmes, and K. Iobst. Processing in memory: the terasys massively parallel PIM array. Computer, 28(4):23--31, 1995. Google ScholarDigital Library
- T. Kgil and T. Mudge. FlashCache: a NAND Flash memory file cache for low power web servers. CASES, 2006. Google ScholarDigital Library
- T. Kgil, D. Roberts, and T. Mudge. Improving nand Flash based disk caches. ISCA, 2008. Google ScholarDigital Library
- B. C. Lee, et al. Architecting phase change memory as a scalable dram alternative. ISCA, 2009. Google ScholarDigital Library
- D. Lewis and H. Lee. Architectural evaluation of 3D stacked RRAM caches. IEEE 3D System Integration Conf., 2009.Google Scholar
- S. Li, et al. McPAT: An integrated power, area and timing modeling framework for multicore and manycore architectures. MICRO, 2009. Google ScholarDigital Library
- K. Lim, et al. Understanding and designing new server architectures for emerging warehouse-computing environments. ISCA, 2008. Google ScholarDigital Library
- D. Patterson, et al.A case for intelligent RAM. IEEE Micro, 1997. Google ScholarDigital Library
- M. K. Qureshi, et al. Enhancing lifetime and security of pcm-based main memory with start-gap wear leveling. MICRO-42, 2009. Google ScholarDigital Library
- M. Qureshi, et al. Scalable high performance main memory system using phase-change memory technology. ISCA, 2009. Google ScholarDigital Library
- V. Reddi, et al. Web Search Using Small Cores: Quantifying the Price of Efficiency. ISCA, 2010.Google Scholar
- E. Riedel, et al. Active disks for large-scale data processing. IEEE Computer, vol 34, , 2001. Google ScholarDigital Library
- S. Rivoire, et al. JouleSort: a balanced energy-efficiency benchmark. SIGMOD, 2007. Google ScholarDigital Library
- P. Clark, et al. Processing Petabytes per Second with the ATLAS Experiment at the LHC in CERN. GPU Tech. Conf., 2010.Google Scholar
- Zichen Xu, et al. Exploring power-performance tradeoffs in database systems. ICDE, 2010.Google Scholar
- Fan Yang, et al. Formalizing mapreduce with CSP. ECBS, 2010. Google ScholarDigital Library
- B. Zhai, et al. Energy efficient near-threshold chip multi-processing. ISLPED, 2007. Google ScholarDigital Library
- W. Zhang and T. Li. Exploring phase change memory and 3D die-stacking for power/thermal friendly, fast and durable memory architectures. PACT, 2009. Google ScholarDigital Library
- P. Zhou, et al. A durable and energy efficient main memory using phase change memory technology. ISCA, 2009. Google ScholarDigital Library
- Y. Zhou et al. Large-scale Parallel Collaborative Filtering for the Netflix Prize. Algo. Aspects in Information and Management, 2008. Google ScholarDigital Library
- D. Abts et al. Energy proportional datacenter networks. ISCA, 2010. Google ScholarDigital Library
- J. Condit et al, Better I/O through byte-addressable, persistent memory. SOSP, 2009. Google ScholarDigital Library
- S. Venkataraman et al. Consistent and Durable Data Structures for Non-Volatile Byte-Addressable Memory. FAST, 2011. Google ScholarDigital Library
- J. Coburn et al. NV-Heaps: Making Persistent Objects Fast and Safe with Next-Generation, Non-Volatile Memories. ASPLOS, 2011. Google ScholarDigital Library
- D. Stukov, G. Snider, D. Steward, and R. Williams. The missing memristor found. Nature, volume 453, pages 80--83, 2008.Google Scholar
Index Terms
- A limits study of benefits from nanostore-based future data-centric system architectures
Recommendations
Mellow writes: extending lifetime in resistive memories through selective slow write backs
ISCA'16Emerging resistive memory technologies, such as PCRAM and ReRAM, have been proposed as promising replacements for DRAM-based main memory, due to their better scalability, low standby power, and non-volatility. However, limited write endurance is a major ...
NVM duet: unified working memory and persistent store architecture
ASPLOS '14: Proceedings of the 19th international conference on Architectural support for programming languages and operating systemsEmerging non-volatile memory (NVM) technologies have gained a lot of attention recently. The byte-addressability and high density of NVM enable computer architects to build large-scale main memory systems. NVM has also been shown to be a promising ...
Reducing write activities on non-volatile memories in embedded CMPs via data migration and recomputation
DAC '10: Proceedings of the 47th Design Automation ConferenceRecent 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, ...
Comments