Marco Vanneschi

k-ary n-trees: high performance networks for massively parallel architectures

The past few years have seen a rise in popularity of massively parallel architectures that use fat-trees as their interconnection networks. In this paper we study the communication performance of a parametric family of fat-trees, the k-ary n-trees, built with constant arity switches interconnected in a regular topology. Through simulation on a 4-ary 4-tree with 256 nodes, we analyze some variants of an adaptive algorithm that utilize wormhole routing with one, two and four virtual channels. The experimental results show that the uniform, bit reversal and transpose traffic patterns are very sensitive to the flow control strategy. In all these cases, the saturation points are between 35-40% of the network capacity with one virtual channel, 55-60% with two virtual channels and around 75% with four virtual channels. The complement traffic, a representative of the class of the congestion-free communication patterns, reaches an optimal performance, with a saturation point at 97% of the capacity for all flow control strategies.

A methodology for the development and the support of massively parallel programs

Abstract The most important features that a parallel programming language should provide are portability, modularity , and ease of use , as well as performance and efficiency . Current parallel languages are only characterized by some of these features. For instance, most of these languages allow programmers to efficiently exploit the massively parallel target machine. Unfortunately, the estimation of the performance of each application is usually made by the programmer, without the support of any tool. Moreover, the programs produced by using such languages are not portable or easily modifiable. Here, we present a methodology to easily write efficient, high performance and portable massively parallel programs. The methodology is based on the definition of a new explicitly parallel programming language, namely P 3 L , and of a set of compiling tools that perform automatic adaptation of the program features to the target architecture hardware. Target architectures taken into account here are general purpose, distributed memory, MIMD architectures. These architectures provide the scalability and low cost features that are necessary to tackle the goal of massively parallel computing. Following the P 3 L methodology, the programmer has just to specify the kind of parallelism he is going to exploit (pipeline, farm, data, etc.) in the parallel application. Then, P 3 L programming tools automatically generate the process network that implements and optimizes, for the given target architecture, the particular kind of parallelism the programmer indicated as the most suitable for the application.

SkIE: a heterogeneous environment for HPC applications

Abstract Technological directions for innovative HPC software environments are discussed in this paper. We focus on industrial user requirements of heterogeneous multidisciplinary applications, performance portability, rapid prototyping and software reuse, integration and interoperability of standard tools. The various issues are demonstrated with reference to the PQE2000project and its programming environment Skeleton-based Integrated Environment ( SkIE ). SkIE includes a coordination language, SkIECL , allowing the designers to express, in a primitive and structured way, efficient combinations of data parallelism and task parallelism. The goal is achieving fast development and good efficiency for applications in different areas. Modules developed with standard languages and tools are encapsulated into SkIECL structures to form the global application. Performance models associated to the coordination language allow powerful optimizations to be introduced both at run time and at compile time without the direct intervention of the programmer. The paper also discusses the features of the SkIE environment related to debugging, performance analysis tools, visualization and graphical user interface. A discussion of the results achieved in some applications developed using the environment concludes the paper.

ASSIST as a research framework for high-performance grid programming environments

The research activity of our group at the Department of Computer Science, University of Pisa, is focused on programming models and environments for the development of high-performance multidisciplinary applications. The enabling computing platforms we are considering are complex distributed architectures, whose nodes are parallel machines of any kind, including PC/workstation clusters. In general such platforms are characterized by heterogeneity of nodes, and by dynamicity in resource management and allocation. In this context, Grid platforms at various levels of integration [25] are of main interest, including complex distributed structures of general and dedicated subsystems, private heterogeneous networks, and systems for pervasive and ubiquitous computing. In the following, we shall speak of Grids to refer to such architectural scenario.

Dynamic reconfiguration of grid-aware applications in ASSIST

Current grid-aware applications are implemented on top of low-level libraries by developers who are experts on grid middleware architecture. This approach can hardly support the additional complexity of QoS control in real applications. We discuss a novel approach used in the ASSIST programming environment to implement/guarantee user provided QoS contracts in a transparent and effective way. Our approach is based on the implementation of automatic run-time reconfiguration of ASSIST application executions triggered by mismatch between the user provided QoS contract and the actual performance values achieved.

The Implementation of ASSIST, an Environment for Parallel and Distributed Programming

We describe the implementation of ASSIST, a programming environment for parallel and distributed programs. Its coordination language is based of the parallel skeleton model, extended with new features to enhance expressiveness, parallel software reuse, software component integration and interfacing to external resources. The compilation process and the structure of the run-time support of ASSIST are discussed with respect to the issues introduced by the new characteristics, presenting an analysis of the first test results.

Components for High-Performance Grid Programming in Grid.IT

This chapter presents the main ideas of the high-performance component-based Grid programming environment of the Grid.it project. High-performance components are characterized by a programming model that integrates the concepts of structured parallelism, component interaction, compositionality, and adaptivity. We show that ASSIST, the prototype of parallel programming environment currently under development at our group, is a suitable basis to capture all the desired features of the component model in a flexible and efficient manner. For the sake of interoperability, ASSIST modules or programs are automatically encapsulated in standard frameworks; currently, we are experimenting Web Services and the CORBA Component Model. Grid applications, built as compositions of ASSIST components and possibly other existing (legacy) components, are supported by an innovative Grid Abstract Machine, that includes essential abstractions of standard middleware services and a hierarchical Application Manager (AM). AM supports static allocation and dynamic reallocation of adaptive applications according to a performance contract, a reconfiguration strategy, and a performance model.

High-performance data mining with skeleton-based structured parallel programming

We show how to apply a Structured Parallel Programming methodology based on skeletons to Data Mining problems, reporting several results about three commonly used mining techniques, namely association rules, decision tree induction and spatial clustering. We analyze the structural patterns common to these applications, looking at application performance and software engineering efficiency. Our aim is to clearly state what features a Structured Parallel Programming Environment should have to be useful for parallel Data Mining. Within the skeleton-based PPE SkIE that we have developed, we study the different patterns of data access of parallel implementations of Apriori, C4.5 and DBSCAN. We need to address large partitions reads, frequent and sparse access to small blocks, as well as an irregular mix of small and large transfers, to allow efficient development of applications on huge databases. We examine the addition of an object/component interface to the skeleton structured model, to simplify the development of environment-integrated, parallel Data Mining applications.

Dynamicity in distributed applications: issues, problems and the ASSIST approach

Abstract Dynamic reconfiguration refers to the ability of changing properties and structure of a distributed system, while it is running. It is essential when designing applications that need to adapt to unpredictable events, run on non-dedicated environments; for which configuration cannot be statically determined, or may change at runtime. Grid computing also gave new emphasis to the topic, being based on dynamic environments by definition. The first part of this work tries to give a definite organization to known literature and existing experiments, providing a general overview of the problem. Then we describe our approach to dynamicity in the context of the ASSIST programming environment and show how it is able to provide users with many of the required abstractions to develop adaptive, high-performance, distributed applications.

PQE2000: HPC tools for industrial applications

It is time for high-quality products to emerge for HPC systems, but achieving this goal requires a new way of transferring results from the research environment to the application marketplace. As the product of three Italian research agencies and one industrial company: the National Research Council (CNR), National Agency for New Technologies, Energy, and Environment (ENEA), National Institute for Nuclear Physics (NFN), and Finmeccanicas Quadrics Supercomputers World Ltd. (QSW), the PQE2000 project adopts this viewpoint with respect to innovative software technologies and composite MPP architectures for petaflops computing. The author discusses the PQE2000 software technology, shared memory implementations and the PQE2000 architecture for MPP computations.