MVA-Based Probabilistic Model of Shared Memory with a Round Robin Arbiter for Predicting Performance with Heterogeneous Workload

Memory access contention can be a cause of performance problems and should be assessed at early stages of development. We devised a probabilistic model of shared memory for performance estimation. The calculation time is polynomial in the number of processors. The model is applicable for the region of high and heterogeneous bandwidth utilization. A round-robin arbiter is modeled using Mean Value Analysis MVA based approximations and incorporating non-linear dependence to the bandwidth utilization. To evaluate our model, estimated execution time is compared with the measured execution time of benchmark programs with memory access contention. We find a maximum error of 4.2% for the round-robin arbitration when we compensate for the burstiness of accesses.

[1]  Steffen Zschaler,et al.  Modular DSLs for Flexible Analysis: An e-Motions Reimplementation of Palladio , 2014, ECMFA.

[2]  Frederica Darema,et al.  Memory access patterns of parallel scientific programs , 1987, SIGMETRICS '87.

[3]  Trevor Mudge,et al.  MiBench: A free, commercially representative embedded benchmark suite , 2001 .

[4]  Tao Li,et al.  Using Analytical Models to Efficiently Explore Hardware Transactional Memory and Multi-Core Co-Design , 2008, 2008 20th International Symposium on Computer Architecture and High Performance Computing.

[5]  Yoshifumi Sakamoto,et al.  A Model-Based Method for Evaluating Embedded System Performance by Abstraction of Execution Traces , 2010, ECMFA.

[6]  Henk Jonkers,et al.  A probabilistic approach to parallel system performance modelling , 1995, Proceedings of the Twenty-Eighth Annual Hawaii International Conference on System Sciences.

[7]  John P. Hayes,et al.  Analysis of Multiple-Bus Interconnection Networks , 1986, J. Parallel Distributed Comput..

[8]  Mary K. Vernon,et al.  Analytic Evaluation of Shared-Memory Architectures , 2003, IEEE Trans. Parallel Distributed Syst..

[9]  David A. Patterson,et al.  Computer Architecture, Fifth Edition: A Quantitative Approach , 2011 .

[10]  Bohdan Smilauer.  General Model for Memory Interference in Multiprocessors and Mean Value Analysis , 1985, IEEE Transactions on Computers.

[11]  CORNELIS H. HOOGENDOORN.  A General Model for Memory Interference in Multiprocessors , 1977, IEEE Transactions on Computers.

[12]  Vittorio Cortellessa,et al.  Integrating Software Models and Platform Models for Performance Analysis , 2007, IEEE Transactions on Software Engineering.

[13]  Donald E. Thomas,et al.  Stochastic Contention Level Simulation for Single-Chip Heterogeneous Multiprocessors , 2010, IEEE Transactions on Computers.

[14]  Fernando Gustavo Tinetti,et al.  Computer Architecture: A Quantitative Approach J. L. Hennessy, D. A. Patterson Morgan Kaufman, 4th Edition, 2007 , 2008 .

[15]  Yoshifumi Sakamoto,et al.  Coarse-grained simulation method for performance evaluation a of shared memory system , 2011, 16th Asia and South Pacific Design Automation Conference (ASP-DAC 2011).

[16]  Matthias Gries,et al.  Methods for evaluating and covering the design space during early design development , 2004, Integr..