Simple Performance Bounds for Multicore and Parallel Channel Systems

A simple modification of existing divisible load scheduling algorithms, boosting link speed by M for M parallel channels per link, allows time optimal load scheduling and performance prediction for parallel channel systems. The situation for multicore models is more complex but can be handled by a substitution involving equivalent processor speed. These modifications yield upper bounds on such parallel systems' performance. This concept is illustrated for ideal single level (star) tree networks under a variety of scheduling policies. Less than ideal parallelism can also be modeled though mechanisms of inefficiency require further research.

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

[2]  Nathan R. Tallent,et al.  Identifying Performance Bottlenecks in Work-Stealing Computations , 2009, Computer.

[3]  Henri Casanova,et al.  UMR: a multi-round algorithm for scheduling divisible workloads , 2003, Proceedings International Parallel and Distributed Processing Symposium.

[4]  Thomas G. Robertazzi,et al.  Optimal Divisible Load Scheduling and Markov Chain Models , 2003 .

[5]  Bharadwaj Veeravalli,et al.  Scheduling Multisource Divisible Loads on Arbitrary Networks , 2010, IEEE Transactions on Parallel and Distributed Systems.

[6]  Thomas G. Robertazzi,et al.  Bus-oriented load sharing for a network of sensor driven processors , 1991, IEEE Trans. Syst. Man Cybern..

[7]  William J. Dally,et al.  Research Challenges for On-Chip Interconnection Networks , 2007, IEEE Micro.

[8]  C. Siva Ram Murthy,et al.  Distributed computation for a hypercube network of sensor-driven processors with communication delays including setup time , 1998, IEEE Trans. Syst. Man Cybern. Part A.

[9]  Thomas G. Robertazzi.  A Product Form Solution for Tree Networks with Divisible Loads , 2011, Parallel Process. Lett..

[10]  Maciej Drozdowski,et al.  THE PERFORMANCE LIMITS OF A TWO-DIMENSIONAL NETWORK OF LOAD-SHARING PROCESSORS , 2011 .

[11]  Henri Casanova,et al.  Parallel Algorithms , 2019, Design and Analysis of Algorithms.

[12]  Henri Casanova,et al.  A realistic network/application model for scheduling divisible loads on large-scale platforms , 2005, 19th IEEE International Parallel and Distributed Processing Symposium.

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

[14]  Debasish Ghose,et al.  Multi-installment load distribution in tree networks with delays , 1995 .

[15]  Jacek Blazewicz,et al.  Scheduling Divisible Jobs on Hypercubes , 1995, Parallel Comput..

[16]  Debasish Ghose,et al.  Divisible Load Theory: A New Paradigm for Load Scheduling in Distributed Systems , 2004, Cluster Computing.

[17]  Thomas G. Robertazzi.  Processor equivalence for a linear daisy chain of load sharing processors , 1992 .

[18]  T.G. Robertazzi,et al.  Optimal divisible job load sharing for bus networks , 1996, IEEE Transactions on Aerospace and Electronic Systems.

[19]  Jacek Blazewicz,et al.  Distributed Processing of Divisible Jobs with Communication Startup Costs , 1997, Discret. Appl. Math..

[20]  Yves Robert,et al.  Optimal algorithms for scheduling divisible workloads on heterogeneous systems , 2003, Proceedings International Parallel and Distributed Processing Symposium.

[21]  Arnold L. Rosenberg,et al.  Sharing partitionable workloads in heterogeneous NOWs: greedier is not better , 2001, Proceedings 42nd IEEE Symposium on Foundations of Computer Science.

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

[23]  Thomas G. Robertazzi,et al.  Grid scheduling divisible loads from two sources , 2009, Comput. Math. Appl..

[24]  Thomas G. Robertazzi,et al.  Distributed computation for a tree network with communication delays , 1990 .

[25]  Debasish Ghose,et al.  Distributed Computation with Communication Delays: Asymptotic Performance Analysis , 1994, J. Parallel Distributed Comput..

[26]  Thomas G. Robertazzi.  Processor equivalence for daisy chain load sharing processors , 1993 .

[27]  M. Moges,et al.  Wireless sensor networks: scheduling for measurement and data reporting , 2006, IEEE Transactions on Aerospace and Electronic Systems.

[28]  Xian-He Sun,et al.  Reevaluating Amdahl's law in the multicore era , 2010, J. Parallel Distributed Comput..

[29]  Thomas G. Robertazzi,et al.  Scalable Scheduling for Clusters and Grids Using Cut Through Switching , 2004 .

[30]  Theron Lammie,et al.  A Linear Daisy Chain with Two Divisible Load Sources , 2005 .

[31]  T.G. Robertazzi,et al.  Multi-Source Grid Scheduling for Divisible Loads , 2006, 2006 40th Annual Conference on Information Sciences and Systems.

[32]  Thomas G. Robertazzi.  Networks and grids - technology and theory , 2007, Information technology.

[33]  Dantong Yu,et al.  Data Intensive Grid Scheduling: Multiple Sources with Capacity Constraints , 2003 .

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

[35]  Mark D. Hill,et al.  Amdahl's Law in the Multicore Era , 2008 .