Abstract
Utilizing analytical models to evaluate proposals or provide guidance in high-level architecture decisions is been becoming more and more attractive. A certain number of methods have emerged regarding cache behaviors and quantified insights in the last decade, such as the stack distance theory and the memory level parallelism (MLP) estimations. However, prior research normally oversimplified the factors that need to be considered in out-of-order processors, such as the effects triggered by reordered memory instructions, and multiple dependences among memory instructions, along with the merged accesses in the same MSHR entry. These ignored influences actually result in low and unstable precisions of recent analytical models.
By quantifying the aforementioned effects, this article proposes a cache performance evaluation framework equipped with three analytical models, which can more accurately predict cache misses, MLPs, and the average cache miss service time, respectively. Similar to prior studies, these analytical models are all fed with profiled software characteristics in which case the architecture evaluation process can be accelerated significantly when compared with cycle-accurate simulations.
We evaluate the accuracy of proposed models compared with gem5 cycle-accurate simulations with 16 benchmarks chosen from Mobybench Suite 2.0, Mibench 1.0, and Mediabench II. The average root mean square errors for predicting cache misses, MLPs, and the average cache miss service time are around 4%, 5%, and 8%, respectively. Meanwhile, the average error of predicting the stall time due to cache misses by our framework is as low as 8%. The whole cache performance estimation can be sped by about 15 times versus gem5 cycle-accurate simulations and 4 times when compared with recent studies. Furthermore, we have shown and studied the insights between different performance metrics and the reorder buffer sizes by using our models. As an application case of the framework, we also demonstrate how to use our framework combined with McPAT to find out Pareto optimal configurations for cache design space explorations.
- Jeffrey M. Abramson, David B. Papworth, Haitham H. Akkary, Andrew F. Glew, Glenn J. Hinton, Kris G. Konigsfeld, Paul D. Madland, et al. 1998. Out-of-order processor with a memory subsystem which handles speculatively dispatched load operations. (May 12, 1998). US Patent 5,751,983.Google Scholar
- Kapil Anand and Rajeev Barua. 2015. Instruction-cache locking for improving embedded systems performance. ACM Transactions on Embedded Computing Systems 14, 3 (2015), 53. Google ScholarDigital Library
- Fabrice Bellard. 2005. QEMU, a fast and portable dynamic translator. In USENIX Annual Technical Conference, FREENIX Track. 41--46. Google ScholarDigital Library
- Erik Berg and Erik Hagersten. 2004. StatCache: A probabilistic approach to efficient and accurate data locality analysis. In Proceedings of the 2004 IEEE International Symposium on ISPASS Performance Analysis of Systems and Software. IEEE, 20--27. Google ScholarDigital Library
- Erik Berg, Håkan Zeffer, and Erik Hagersten. 2006. A statistical multiprocessor cache model. In Proceedings of the 2006 IEEE International Symposium on Performance Analysis of Systems and Software. IEEE, 89--99.Google ScholarCross Ref
- Nathan Binkert, Bradford Beckmann, Gabriel Black, Steven K. Reinhardt, Ali Saidi, Arkaprava Basu, Joel Hestness, Derek R. Hower, Tushar Krishna, Somayeh Sardashti, et al. 2011. The gem5 simulator. ACM SIGARCH Computer Architecture News 39, 2 (2011), 1--7. Google ScholarDigital Library
- J. Bolaria. 2012. Cortex-A57 extends ARM’s reach. Microprocessor Report 11, 5 (2012), 12--1.Google Scholar
- Maximilien Breughe, Stijn Eyerman, and Lieven Eeckhout. 2012. A mechanistic performance model for superscalar in-order processors. In Proceedings of the 2012 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS’12). IEEE, 14--24. Google ScholarDigital Library
- Calin CaBcaval and David A. Padua. 2003. Estimating cache misses and locality using stack distances. In Proceedings of the 17th Annual International Conference on Supercomputing. ACM, 150--159. Google ScholarDigital Library
- Jian Chen, Lizy Kurian John, and Dimitris Kaseridis. 2011. Modeling program resource demand using inherent program characteristics. In Proceedings of the ACM SIGMETRICS Joint International Conference on Measurement and Modeling of Computer Systems. ACM, 1--12. Google ScholarDigital Library
- Tien-Fu Chen and Jean-Loup Baer. 1992. Reducing Memory Latency via Non-Blocking and Prefetching Caches, Vol. 27. ACM. Google ScholarDigital Library
- Yuan Chou, Brian Fahs, and Santosh Abraham. 2004. Microarchitecture optimizations for exploiting memory-level parallelism. In ACM SIGARCH Computer Architecture News, Vol. 32. IEEE Computer Society, 76. Google ScholarDigital Library
- Gyanesh Das, Prasant Kumar Pattnaik, and Sasmita Kumari Padhy. 2014. Artificial neural network trained by particle swarm optimization for non-linear channel equalization. Expert Systems with Applications 41, 7 (2014), 3491--3496. Google ScholarDigital Library
- Roeland J. Douma, Sebastian Altmeyer, and Andy D. Pimentel. 2015. Fast and precise cache performance estimation for out-of-order execution. In Design, Automation 8 Test in Europe Conference & Exhibition (DATE’15). IEEE, 1132--1137. Google ScholarDigital Library
- Jan Draisma, Emil Horobeţ, Giorgio Ottaviani, Bernd Sturmfels, and Rekha Thomas. 2014. The Euclidean distance degree. In Proceedings of the 2014 Symposium on Symbolic-Numeric Computation. ACM, 9--16. Google ScholarDigital Library
- David Eklov and Erik Hagersten. 2010. StatStack: Efficient modeling of LRU caches. In Proceedings of the 2010 IEEE International Symposium on Performance Analysis of Systems 8 Software (ISPASS’10). IEEE, 55--65.Google ScholarCross Ref
- Stijn Eyerman, Lieven Eeckhout, Tejas Karkhanis, and James E. Smith. 2009. A mechanistic performance model for superscalar out-of-order processors. ACM Transactions on Computer Systems 27, 2 (2009), 3. Google ScholarDigital Library
- Stijn Eyerman, Kenneth Hoste, and Lieven Eeckhout. 2011. Mechanistic-empirical processor performance modeling for constructing CPI stacks on real hardware. In Proceedings of the 2011 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS’11). IEEE, 216--226. Google ScholarDigital Library
- Peter Greenhalgh. 2011. Big.LITTLE processing with ARM Cortex-A15 8 Cortex-A7. ARM White Paper (2011), 1--8.Google Scholar
- Matthew R. Guthaus, Jeffrey S. Ringenberg, Dan Ernst, Todd M. Austin, Trevor Mudge, and Richard B. Brown. 2001. MiBench: A free, commercially representative embedded benchmark suite. In Proceedings of the 2001 IEEE International Workshop on Workload Characterization (WWC-4’01). IEEE, 3--14. Google ScholarDigital Library
- Anthony Gutierrez, Joseph Pusdesris, Ronald G. Dreslinski, Trevor Mudge, Chander Sudanthi, Christopher D. Emmons, Mitchell Hayenga, and Nigel Paver. 2014. Sources of error in full-system simulation. In Proceedings of the 2014 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS’14). IEEE, 13--22.Google ScholarCross Ref
- Linley Gwennap. 1997. Intel, HP make EPIC disclosure. Microprocessor Report 11, 14 (1997), 1--9.Google Scholar
- John A. Hartigan and Manchek A. Wong. 1979. Algorithm AS 136: A k-means clustering algorithm. Journal of the Royal Statistical Society. Series C (Applied Statistics) 28, 1 (1979), 100--108.Google ScholarDigital Library
- John L. Hennessy and David A. Patterson. 2011. Computer Architecture: A Quantitative Approach. Elsevier. Google ScholarDigital Library
- Glenn Hinton, Dave Sager, Mike Upton, Darrell Boggs, et al. 2001. The microarchitecture of the Pentium® 4 processor. In Intel Technology Journal. Citeseer.Google Scholar
- Yongbing Huang, Zhongbin Zha, Mingyu Chen, and Lixin Zhang. 2014. Moby: A mobile benchmark suite for architectural simulators. In Proceedings of the 2014 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS’14). IEEE, 45--54.Google ScholarCross Ref
- Kazuhisa Ishizaka and Takashi Miyazaki. 2013. Cache memory, including miss status/information and a method using the same. (Oct. 8, 2013). US Patent 8,555,001.Google Scholar
- J. Jaleel and Bruce Jacob. 2005. Using virtual load/store queues (VLSQs) to reduce the negative effects of reordered memory instructions. In Proceedings of the 11th International Symposium on High-Performance Computer Architecture, 2005 (HPCA’11). IEEE, 191--200. Google ScholarDigital Library
- Kecheng Ji, Ming Ling, Qin Wang, Longxing Shi, and Jianping Pan. 2017a. AFEC: An analytical framework for evaluating cache performance in out-of-order processors. In Proceedings of the 2017 Design, Automation 8 Test in Europe Conference 8 Exhibition (DATE’17). IEEE, 55--60. Google ScholarDigital Library
- Kecheng Ji, Ming Ling, Yang Zhang, and Longxing Shi. 2017b. An artificial neural network model of LRU-cache misses on out-of-order embedded processors. Microprocessors and Microsystems 50 (2017), 66--79.Google ScholarCross Ref
- Wen Jin, Zhao Jia Li, Luo Si Wei, and Han Zhen. 2000. The improvements of BP neural network learning algorithm. In Proceedings of the 5th International Conference on Signal Processing (WCCC-ICSP’00), Vol. 3. IEEE, 1647--1649.Google Scholar
- Tejas S. Karkhanis and James E. Smith. 2004. A first-order superscalar processor model. In Proceedings of the 31st Annual International Symposium on Computer Architecture, 2004. Proceedings. IEEE, 338--349. Google ScholarDigital Library
- Chunho Lee, Miodrag Potkonjak, and William H. Mangione-Smith. 1997. MediaBench: A tool for evaluating and synthesizing multimedia and communicatons systems. In Proceedings of the 30th Annual ACM/IEEE International Symposium on Microarchitecture. IEEE Computer Society, 330--335. Google ScholarDigital Library
- Sheng Li, Jung Ho Ahn, Richard D. Strong, Jay B. Brockman, Dean M. Tullsen, and Norman P. Jouppi. 2013. The McPAT framework for multicore and manycore architectures: Simultaneously modeling power, area, and timing. ACM Transactions on Architecture and Code Optimization 10, 1 (2013), 5. Google ScholarDigital Library
- Yun Liang and Tulika Mitra. 2008. Cache modeling in probabilistic execution time analysis. In Proceedings of the 45th Annual Design Automation Conference. ACM, 319--324. Google ScholarDigital Library
- Yun Liang and Tulika Mitra. 2013. An analytical approach for fast and accurate design space exploration of instruction caches. ACM Transactions on Embedded Computing Systems 13, 3 (2013), 43. Google ScholarDigital Library
- Chi-Keung Luk, Robert Cohn, Robert Muth, Harish Patil, Artur Klauser, Geoff Lowney, Steven Wallace, Vijay Janapa Reddi, and Kim Hazelwood. 2005. Pin: Building customized program analysis tools with dynamic instrumentation. In ACM Sigplan Notices, Vol. 40. ACM, 190--200. Google ScholarDigital Library
- Sparsh Mittal. 2016. A survey of recent prefetching techniques for processor caches. ACM Computing Surveys 49, 2 (2016), 35. Google ScholarDigital Library
- Onur Mutlu, Jared Stark, Chris Wilkerson, and Yale N. Patt. 2003. Runahead execution: An alternative to very large instruction windows for out-of-order processors. In Proceedings of the 9th International Symposium on High-Performance Computer Architecture (HPCA-9 2003). IEEE, 129--140. Google ScholarDigital Library
- Xiaoyue Pan and Bengt Jonsson. 2015. A modeling framework for reuse distance-based estimation of cache performance. In Proceedings of the 2015 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS’15). IEEE, 62--71.Google ScholarCross Ref
- Lutz Prechelt et al. 1994. Proben1: A set of neural network benchmark problems and benchmarking rules. (1994).Google Scholar
- K. Skadron, P. S. Ahuja, M. Martonosi, and D. W. Clark. 1999. Branch prediction, instruction-window size, and cache size: Performance trade-offs and simulation techniques. IEEE Transactions on Computers 48, 11 (1999), 1260--1281. Google ScholarDigital Library
- Xian-He Sun and Dawei Wang. 2014. Concurrent average memory access time. Computer 47, 5 (2014), 74--80. Google ScholarDigital Library
- Prathap Kumar Valsan, Heechul Yun, and Farzad Farshchi. 2016. Taming non-blocking caches to improve isolation in multicore real-time systems. In Proceedings of the 2016 IEEE Real-Time and Embedded Technology and Applications Symposium (RTAS’16). IEEE, 1--12.Google ScholarCross Ref
- Sam Van den Steen, Stijn Eyerman, Sander De Pestel, Moncef Mechri, Trevor E. Carlson, David Black-Schaffer, Erik Hagersten, and Lieven Eeckhout. 2016. Analytical processor performance and power modeling using micro-architecture independent characteristics. IEEE Transactions on Computers 65, 12 (2016), 3537--3551. Google ScholarDigital Library
- Wei Wang and Tanima Dey. 2011. A survey on ARM Cortex a processors. Retrieved March 2011 from http://www.cs.virginia.edu/shadron/cs8535s11/armcotex.pdf.Google Scholar
- Zhonglei Wang and Jörg Henkel. 2013. Fast and accurate cache modeling in source-level simulation of embedded software. In Proceedings of the Conference on Design, Automation and Test in Europe. EDA Consortium, 587--592. Google ScholarDigital Library
Index Terms
- An Analytical Cache Performance Evaluation Framework for Embedded Out-of-Order Processors Using Software Characteristics
Recommendations
Counter-Based Cache Replacement and Bypassing Algorithms
Recent studies have shown that in highly associative caches, the performance gap between the Least Recently Used (LRU) and the theoretical optimal replacement algorithms is large, motivating the design of alternative replacement algorithms to improve ...
A low-power cache scheme for embedded computing
Issues in embedded single-chip multicore architecturesThis paper proposes an efficient cache scheme to reduce power consumption and conflict misses for single-core or multi-core embedded computing architecture. The proposed cache requires an additional gate stage before it accesses the cache line, which ...
Locating cache performance bottlenecks using data profiling
EuroSys '10: Proceedings of the 5th European conference on Computer systemsEffective use of CPU data caches is critical to good performance, but poor cache use patterns are often hard to spot using existing execution profiling tools. Typical profilers attribute costs to specific code locations. The costs due to frequent cache ...
Comments