Isoefficiency Maps for Divisible Computations in Hierarchical Memory Systems

In this paper we analyze impact of memory hierarchy on divisible load processing. Current computer systems have hierarchical memory systems. The core memory is fast but small, the external memory is large but slow. It is possible to avoid using external memory by parallelizing computations or by processing smaller work chunks sequentially. We will analyze how a combination of these two options improves efficiency of the computations. For this purpose divisible load theory representing data-parallel applications is used. A mathematical model for scheduling divisible computations is formulated as a mixed integer linear program. The model allows for executing multiple load installments sequentially or in parallel. The efficiency of the schedule is analyzed with respect to the impact of load size and machines number. The results are visualized as isoefficiency maps.

[1]  Maciej Drozdowski,et al.  Time and Energy Performance of Parallel Systems with Hierarchical Memory , 2015, Journal of Grid Computing.

[2]  Debasish Ghose,et al.  Scheduling Divisible Loads in Parallel and Distributed Systems , 1996 .

[3]  Selim G. Akl,et al.  Design and analysis of parallel algorithms , 1985 .

[4]  Bharadwaj Veeravalli,et al.  Divisible load scheduling on single-level tree networks with buffer constraints , 2000, IEEE Trans. Aerosp. Electron. Syst..

[5]  Thomas G. Robertazzi,et al.  Ten Reasons to Use Divisible Load Theory , 2003, Computer.

[6]  Pawel Wolniewicz,et al.  Out-of-Core Divisible Load Processing , 2003, IEEE Trans. Parallel Distributed Syst..

[7]  Joanna Berlinska.  Communication scheduling in data gathering networks with limited memory , 2014, Appl. Math. Comput..

[8]  Youmin Zhang,et al.  Numerically robust implementation of multiple-model algorithms , 1999, IEEE Trans. Aerosp. Electron. Syst..

[9]  Maciej Drozdowski,et al.  Isoefficiency Maps for Divisible Computations , 2010, IEEE Transactions on Parallel and Distributed Systems.

[10]  Maciej Drozdowski,et al.  Energy trade-offs analysis using equal-energy maps , 2014, Future Gener. Comput. Syst..

[11]  Maciej Drozdowski,et al.  Scheduling for Parallel Processing , 2009, Computer Communications and Networks.

[12]  Maciej Drozdowski,et al.  Divisible Loads Scheduling in Hierarchical Memory Systems with Time and Energy Constraints , 2015, PPAM.

[13]  Shamsollah Ghanbari,et al.  Comprehensive Review on Divisible Load Theory: Concepts, Strategies, and Approaches , 2014 .

[14]  H. V. Jagadish,et al.  Partitioning Techniques for Large-Grained Parallelism , 1988, IEEE Trans. Computers.

[15]  Debasish Ghose,et al.  Adaptive divisible load scheduling strategies for workstation clusters with unknown network resources , 2005, IEEE Transactions on Parallel and Distributed Systems.

[16]  Steven Swanson,et al.  Refactor, Reduce, Recycle: Restructuring the I/O Stack for the Future of Storage , 2013, Computer.

[17]  Vipin Kumar,et al.  Performance Properties of Large Scale Parallel Systems , 1993, J. Parallel Distributed Comput..