Fernando Guirado

A New Task Graph Model for Mapping Message Passing Applications

The exploitation of parallelism in a message passing platform implies a previous modeling phase of the parallel application as a task graph, which properly reflects its temporal behavior. In this paper, we analyze the classical task graph models of the literature and their drawbacks when modeling message passing programs with an arbitrary task structure. We define a new task graph model called temporal task interaction graph (TTIG) that integrates the classical models used in the literature. The TTIG allows us to explicitly capture the ability of concurrency of adjacent tasks for applications where adjacent tasks can communicate at any point inside them. A mapping strategy is developed from this model, which minimizes the expected execution time by properly exploiting task parallelism. The effectiveness of this approach has been proved in different experimentation scopes for a wide range of message passing applications.

MIP model scheduling for multi-clusters

Multi-cluster environments are composed of multiple clusters that act collaboratively, thus allowing computational problems that require more resources than those available in a single cluster to be treated. However, the degree of complexity of the scheduling process is greatly increased by the resources heterogeneity and the co-allocation process, which distributes the tasks of parallel jobs across cluster boundaries. In this paper, the authors propose a new MIP model which determines the best scheduling for all the jobs in the queue, identifying their resource allocation and its execution order to minimize the overall makespan. The results show that the proposed technique produces a highly compact scheduling of the jobs, producing better resources utilization and lower overall makespan. This makes the proposed technique especially useful for environments dealing with limited resources and large applications.

A new model for static mapping of parallel applications with task and data parallelism

The efficient mapping of parallel tasks is essential in order to exploit the gain from parallelisation. In this work, we focus on modelling and mapping message-passing applications that are defined by the programmer with an arbitrary interaction pattern among tasks. A new model is proposed, known as TTIG (Temporal Task Interaction Graph), which captures not only computation and communication costs, but also the percentages of concurrency between tasks. From this model, a mapping strategy is developed that minimises expected execution time by properly exploiting task parallelism. The effectiveness of this approach has been proven for a real image-processing application on a cluster of PCs.

High Performance computing improvements on bioinformatics consistency-based multiple sequence alignment tools

BGT method improves the execution time of progressive alignment by 62%.OLM decreases the memory requirements by 75%, aligning up twice more sequences than previous method.MTA improves the alignments accuracy, letting it to regain the quality lost due to the performance improvements. Multiple Sequence Alignment (MSA) is essential for a wide range of applications in Bioinformatics. Traditionally, the alignment accuracy was the main metric used to evaluate the goodness of MSA tools. However, with the growth of sequencing data, other features, such as performance and the capacity to align larger datasets, are gaining strength. To achieve these new requirements, without affecting accuracy, the use of high-performance computing (HPC) resources and techniques is crucial. In this paper, we apply HPC techniques in T-Coffee, one of the more accurate but less scalable MSA tools. We integrate three innovative solutions into T-Coffee: the Balanced Guide Tree to increase the parallelism/performance, the Optimized Library Method with the aim of enhancing the scalability and the Multiple Tree Alignment, which explores different alignments in parallel to improve the accuracy. The results obtained show that the resulting tool, MTA-TCoffee, is able to improve the scalability in both the execution time and also the number of sequences to be aligned. Furthermore, not only is the alignment accuracy not affected by these improvements, as would be expected, but it improves significantly. Finally, we emphasize that the presented methods are not just restricted to T-Coffee, but may be implemented in any other alignment tools that use similar algorithms (progressive alignment, consistency or guide trees).

Exploiting parallelism on progressive alignment methods

Multiple Sequence Alignment (MSA) constitutes an extremely powerful tool for important biological applications such as phylogenetic analysis, identification of conserved motifs and domains and structure prediction. In spite of the improvement in speed and accuracy introduced by MSA programs, the computational requirements for large-scale alignments requires high-performance computing and parallel applications. In this paper we present an improvement to a parallel implementation of T-Coffee, a widely used MSA package. Our approximation resolves the bottleneck of the progressive alignment stage on MSA. This is achieved by increasing the degree of parallelism by balancing the guide tree that drives the progressive alignment process. The experimental results show improvements in execution time of over 68% while maintaining the biological accuracy.

Multiple job co-allocation strategy for heterogeneous multi-cluster systems based on linear programming

Multi-cluster environments are composed of multiple clusters of computers that act collaboratively, and thus allowing computational problems to be treated that require more resources than those available in a single cluster. However, the degree of complexity of the scheduling process is greatly increased by the heterogeneity of resources and co-allocation process, which distributes the tasks of parallel jobs across cluster boundaries.This work presents a new scheduling strategy that allocates multiple jobs from the system queue simultaneously on a heterogeneous multicluster, by applying co-allocation when is necessary. Our strategy is composed by a job selection function and a linear programming model to find the best allocation for multiple jobs. The proposed scheduling technique is shown to reduce the execution times of the parallel jobs and the overall response times by 38% compared with other scheduling techniques in the literature.

Modelling message-passing programs for static mapping

An efficient mapping of a parallel program in the processors is vital for achieving a high performance on a parallel computer. When the structure of the parallel program in terms of its task execution times, task dependencies, and amount communication data, is known a priori, mapping can be accomplished statically at compile time. Mapping algorithms start from a parallel application model and map automatically tasks to processors in order to minimise the execution time of the program. In this paper we discuss the current models used in mapping parallel programs: Task Precedence Graph (TPG), Task Interaction Graph (TIG) and we define a new model called Temporal Task Interaction Graph (TTIG). The contribution of the TTIG is that it enhances these two previous models with the ability to explicitly capture the potential degree of parallel execution between adjacent tasks allowing the development of efficient mapping algorithms. Experimentation had been performed in order to show the effectiveness of TTIG model for a set of graphs. The results are compared with the optimal assignment and the obtained using TIG model and they confirm that using the TTIG model, better assignments can be obtained.

Optimizing Latency under Throughput Requirements for Streaming Applications on Cluster Execution

Parallelism in applications that act on a stream of input data can be exploited with two different approaches, spatial and temporal. In this paper we propose a new task mapping algorithm, called EXPERT, to exploit temporal parallelism efficiently when the streaming application is running in a pipeline fashion. We compare the performance of spatial and temporal approaches, in terms of latency and throughput for a video compression application. The results show that the pipeline execution with the task assignment provided by EXPERT algorithm, significantly overcomes spatial parallelism. Additionally, this temporal parallelism presents better scalability results when the dimension of the problem is augmented

Performance prediction using an application-oriented mapping tool

Simulation frameworks are widely used to carry out performance predictions of parallel programs. In general, these environments do not support the use of automatic mapping mechanisms for assigning tasks to processors. We present a tool called pMAP (predicting the best mapping of parallel applications) that performs the mapping process of message-passing applications starting from the characterization of application behaviour. We show that it is important to explore in these simulation frameworks, not only how many resources are needed to achieve good results, but also how best to map the application onto the parallel system. We study the evaluation of the pMAP tool integrated with the commercial simulator DIMEMAS. The results show that for message-passing applications with different task structures, great improvements in the speedup can be obtained when simulations are carried out with the mappings generated by pMAP.

Improving multiple sequence alignment biological accuracy through genetic algorithms

Accuracy on multiple sequence alignments (MSA) is of great significance for such important biological applications as evolution and phylogenetic analysis, homology and domain structure prediction. In such analyses, alignment accuracy is crucial. In this paper, we investigate a combined scoring function capable of obtaining a good approximation to the biological quality of the alignment. The algorithm uses the information obtained by the different quality scores in order to improve the accuracy. The results show that the combined score is able to evaluate alignments better than the isolated scores.