No abstract

We describe our work on improving the performance of collective communication operations in MPICH for clusters connected by switched networks. For each collective operation, we use multiple algorithms depending on the message size, with the goal of minimizing latency for short messages and minimizing bandwidth use for long messages. Although we have implemented new algorithms for all MPI (Message Passing Interface) collective operations, because of limited space we describe only the algorithms for allgather, broadcast, all-to-all, reduce-scatter, reduce, and allreduce. Performance results on a Myrinet-connected Linux cluster and an IBM SP indicate that, in all cases, the new algorithms significantly outperform the old algorithms used in MPICH on the Myrinet cluster, and, in many cases, they outperform the algorithms used in IBM's MPI on the SP. We also explore in further detail the optimization of two of the most commonly used collective operations, allreduce and reduce, particularly for long messages and nonpower-of-two numbers of processes. The optimized algorithms for these operations perform several times better than the native algorithms on a Myrinet cluster, IBM SP, and Cray T3E. Our results indicate that to achieve the best performance for a collective communication operation, one needs to use a number of different algorithms and select the right algorithm for a particular message size and number of processes.

Over the last twenty years, the open source community has provided more and more software on which the world's High Performance Computing (HPC) systems depend for performance and productivity. The community has invested millions of dollars and years of effort to build key components. But although the investments in these separate software elements have been tremendously valuable, a great deal of productivity has also been lost because of the lack of planning, coordination, and key integration of technologies necessary to make them work together smoothly and efficiently, both within individual PetaScale systems and between different systems. It seems clear that this completely uncoordinated development model will not provide the software needed to support the unprecedented parallelism required for peta/exascale computation on millions of cores, or the flexibility required to exploit new hardware models and features, such as transactional memory, speculative execution, and GPUs. This report describes the work of the community to prepare for the challenges of exascale computing, ultimately combing their efforts in a coordinated International Exascale Software Project.

Over the past few years resilience has became a major issue for high-performance computing (HPC) systems, in particular in the perspective of large petascale systems and future exascale systems. These systems will typically gather from half a million to several millions of central processing unit (CPU) cores running up to a billion threads. From the current knowledge and observations of existing large systems, it is anticipated that exascale systems will experience various kind of faults many times per day. It is also anticipated that the current approach for resilience, which relies on automatic or application level checkpoint/ restart, will not work because the time for checkpointing and restarting will exceed the mean time to failure of a full system. This set of projections leaves the community of fault tolerance for HPC systems with a difficult challenge: finding new approaches, which are possibly radically disruptive, to run applications until their normal termination, despite the essentially unstable nature of exascale systems. Yet, the community has only five to six years to solve the problem. This white paper synthesizes the motivations, observations and research issues considered as determinant of several complimentary experts of HPC in applications, programming models, distributed systems and system management.

No abstract

No abstract

No abstract

-Parallel I/O prefetching is considered to be effective in improving I/O performance. However, the effectiveness depends on determining patterns among future I/O accesses swiftly and fetching data in time, which is difficult to achieve in general. In this study, we propose an I/O signature-based prefetching strategy. The idea is to use a predetermined I/O signature of an application to guide prefetching. To put this idea to work, we first derived a classification of patterns and introduced a simple and effective signature notation to represent patterns. We then developed a toolkit to trace and generate I/O signatures automatically. Finally, we designed and implemented a thread-based client-side collective prefetching cache layer for MPI-IO library to support prefetching. A prefetching thread reads I/O signatures of an application and adjusts them by observing I/O accesses at runtime. Experimental results show that the proposed prefetching method improves I/O performance significantly for applications with complex patterns.