LoGPC: modeling network contention in message-passing programs

In many real applications, for example those with frequent and irregular communication patterns or those using large messages, network contention and contention for message processing resources can be a significant part of the total execution time. This paper presents a new cost model, called LoGPC, that extends the LogP [9] and LogGP [4] models to account for the impact of network contention and network interface DMA behavior on the performance of message-passing programs.We validate LoGPC by analyzing three applications implemented with Active Messages [11, 18] on the MIT Alewife multiprocessor. Our analysis shows that network contention accounts for up to 50% of the total execution time. In addition, we show that the impact of communication locality on the communication costs is at most a factor of two on Alewife. Finally, we use the model to identify tradeoffs between synchronous and asynchronous message passing styles.

[1]  D.E. Culler,et al.  Effects Of Communication Latency, Overhead, And Bandwidth In A Cluster Architecture , 1997, Conference Proceedings. The 24th Annual International Symposium on Computer Architecture.

[2]  Kirk L. Johnson.  The impact of communication locality on large-scale multiprocessor performance , 1992, ISCA '92.

[3]  Anant Agarwal,et al.  Limits on Interconnection Network Performance , 1991, IEEE Trans. Parallel Distributed Syst..

[4]  Andrew A. Chien,et al.  A comparison of architectural support for messaging in the TMC CM-5 and the Cray T3D , 1995, Proceedings 22nd Annual International Symposium on Computer Architecture.

[5]  Andrea C. Arpaci-Dusseau,et al.  Parallel programming in Split-C , 1993, Supercomputing '93. Proceedings.

[6]  Mihalis Yannakakis,et al.  Towards an architecture-independent analysis of parallel algorithms , 1990, STOC '88.

[7]  Albert G. Greenberg.  On the time complexity of broadcast communication schemes (Preliminary Version) , 1982, STOC '82.

[8]  Seth Copen Goldstein,et al.  Active messages: a mechanism for integrating communication and computation , 1998, ISCA '98.

[9]  David A. Patterson,et al.  Logp quantified: the case for low-overhead local area networks , 1995 .

[10]  Peter Thanisch,et al.  Analysis of multicomputer schedules in cost and latency model of communication , 1997 .

[11]  Seth Copen Goldstein,et al.  Active Messages: A Mechanism for Integrated Communication and Computation , 1992, [1992] Proceedings the 19th Annual International Symposium on Computer Architecture.

[12]  Ramesh Subramonian,et al.  LogP: towards a realistic model of parallel computation , 1993, PPOPP '93.

[13]  Chris J. Scheiman,et al.  LogGP: incorporating long messages into the LogP model—one step closer towards a realistic model for parallel computation , 1995, SPAA '95.

[14]  T. von Eicken,et al.  Parallel programming in Split-C , 1993, Supercomputing '93.

[15]  Victor Lee,et al.  Exploiting two-case delivery for fast protected messaging , 1998, Proceedings 1998 Fourth International Symposium on High-Performance Computer Architecture.

[16]  Richard P. Martin,et al.  LogP Performance Assessment of Fast Network Interfaces , 1995 .

[17]  A. Krishnamurthy,et al.  Parallel Programming in Split - , 1993 .

[18]  Andrea C. Arpaci-Dusseau,et al.  Fast Parallel Sorting Under LogP: Experience with the CM-5 , 1996, IEEE Trans. Parallel Distributed Syst..

[19]  Mihalis Yannakakis,et al.  Towards an Architecture-Independent Analysis of Parallel Algorithms , 1990, SIAM J. Comput..

[20]  N. Madsen.  Divergence preserving discrete surface integral methods for Maxwell's curl equations using non-orthogonal unstructured grids , 1995 .

[21]  Marc Snir,et al.  The Performance of Multistage Interconnection Networks for Multiprocessors , 1983, IEEE Transactions on Computers.

[22]  Seth Copen Goldstein,et al.  Active messages: a mechanism for integrating communication and computation , 1998, ISCA '98.

[23]  Ricardo Bianchini,et al.  The MIT Alewife machine: architecture and performance , 1995, Proceedings 22nd Annual International Symposium on Computer Architecture.

[24]  Rajeev Barua,et al.  The sensitivity of communication mechanisms to bandwidth and latency , 1998, Proceedings 1998 Fourth International Symposium on High-Performance Computer Architecture.

[25]  Mary K. Vernon,et al.  LoPC: modeling contention in parallel algorithms , 1997, PPOPP '97.

[26]  Richard P. Martin,et al.  Effects Of Communication Latency, Overhead, And Bandwidth In A Cluster Architecture , 1997, Conference Proceedings. The 24th Annual International Symposium on Computer Architecture.

[27]  Thomas G. Robertazzi,et al.  The Performance of Multistage Interconnection Networks for Multiprocessors , 1993 .

[28]  John L. Hennessy,et al.  The Effects of Latency, Occupancy, and Bandwidth in Distributed Shared Memory Multiprocessors , 1995 .