A buffering layer to support derived types and proprietary networks for Java HPC

MPJ Express is our implementation of MPI-like bindings for Java. In this paper we discuss our intermediate buffering layer that makes use of the so-called direct byte buffers introduced in the Java New I/O package. The purpose of this layer is to support the implementation of derived datatypes. MPJ Express is the first Java messaging library that implements this feature using pure Java. In addition, this buffering layer allows efficient implementation of communication devices based on proprietary networks such as Myrinet. In this paper we evaluate the performance of our buffering layer and demonstrate the usefulness of direct byte buffers. Also, we evaluate the performance of MPJ Express against other messaging systems using Myrinet and show that our buffering layer has made it possible to avoid the overheads suffered by other Java systems such as mpiJava that relies on the Java Native Interface.

A buffering layer to support derived types and proprietary networks for Java HPC

MPJ Express is our implementation of MPI-like bindings for Java. In this paper we discuss our intermediate buffering layer that makes use of the so-called direct byte buffers introduced in the Java New I/O package. The purpose of this layer is to support the implementation of derived datatypes. MPJ Express is the first Java messaging library that implements this feature using pure Java. In addition, this buffering layer allows efficient implementation of communication devices based on proprietary networks such as Myrinet. In this paper we evaluate the performance of our buffering layer and demonstrate the usefulness of direct byte buffers. Also, we evaluate the performance of MPJ Express against other messaging systems using Myrinet and show that our buffering layer has made it possible to avoid the overheads suffered by other Java systems such as mpiJava that relies on the Java Native Interface.

A rare structure at the syntax-discourse interface: heritage and Spanish-dominant native speakers weigh in

The present study examines knowledge of the discourse-appropriateness of Clitic Right Dislocation (CLRD) in a population of Heritage (HS) and Spanish-dominant Native Speakers in order to test the predictions of the Interface Hypothesis (IH; Sorace 2011). The IH predicts that speakers in language contact situations will experience difficulties with integrating information involving the interface of syntax and discourse modules. CLRD relates a dislocated constituent to a discourse antecedent, requiring integration of syntax and pragmatics. Results from an acceptability judgment task did not support the predictions of the IH. No statistical differences between the HSs’ performance and that of L1-dominant native speakers were evidenced when participants were presented with an offline task. Thus, our study did not find any evidence of “incomplete acquisition” (Montrul 2008) as it pertains to this specific linguistic structure.

The adsorbed conformation of globular proteins at the air/water interface

External reflection FTIR spectroscopy and surface pressure measurements were used to compare conformational changes in the adsorbed structures of three globular proteins at the air/water interface. Of the three proteins studied, lysozyme, bovine serum albumin and P-lactoglobulin, lysozyme was unique in its behaviour. Lysozyme adsorption was slow, taking approximately 2.5 h to reach a surface pressure plateau (from a 0.07 mM solution), and led to significant structural change. The FTIR spectra revealed that lysozyme formed a highly networked adsorbed layer of unfolded protein with high antiparallel beta-sheet content and that these changes occurred rapidly (within 10 min). This non-native secondary structure is analogous to that of a 3D heat-set protein gel, suggesting that the adsorbed protein formed a highly networked interfacial layer. Albumin and P-lactoglobulin adsorbed rapidly (reaching a plateau within 10 min) and with little chance to their native secondary structure.

Nested parallelism for multi-core HPC systems using Java

Since its introduction in 1993, the Message Passing Interface (MPI) has become a de facto standard for writing High Performance Computing (HPC) applications on clusters and Massively Parallel Processors (MPPs). The recent emergence of multi-core processor systems presents a new challenge for established parallel programming paradigms, including those based on MPI. This paper presents a new Java messaging system called MPJ Express. Using this system, we exploit multiple levels of parallelism - messaging and threading - to improve application performance on multi-core processors. We refer to our approach as nested parallelism. This MPI-like Java library can support nested parallelism by using Java or Java OpenMP (JOMP) threads within an MPJ Express process. Practicality of this approach is assessed by porting to Java a massively parallel structure formation code from Cosmology called Gadget-2. We introduce nested parallelism in the Java version of the simulation code and report good speed-ups. To the best of our knowledge it is the first time this kind of hybrid parallelism is demonstrated in a high performance Java application. (C) 2009 Elsevier Inc. All rights reserved.

A comparative study of Java and C performance in two large-scale parallel applications

In the 1990s the Message Passing Interface Forum defined MPI bindings for Fortran, C, and C++. With the success of MPI these relatively conservative languages have continued to dominate in the parallel computing community. There are compelling arguments in favour of more modern languages like Java. These include portability, better runtime error checking, modularity, and multi-threading. But these arguments have not converted many HPC programmers, perhaps due to the scarcity of full-scale scientific Java codes, and the lack of evidence for performance competitive with C or Fortran. This paper tries to redress this situation by porting two scientific applications to Java. Both of these applications are parallelized using our thread-safe Java messaging system—MPJ Express. The first application is the Gadget-2 code, which is a massively parallel structure formation code for cosmological simulations. The second application uses the finite-domain time-difference method for simulations in the area of computational electromagnetics. We evaluate and compare the performance of the Java and C versions of these two scientific applications, and demonstrate that the Java codes can achieve performance comparable with legacy applications written in conventional HPC languages. Copyright © 2009 John Wiley & Sons, Ltd.

Amyloid formation: interface influence

The causes of pathological conditions such as Alzheimer’s and Parkinson’s diseases are becoming better understood. Proteins that misfold from their native structure to form aggregates of β-sheet fibrils — termed amyloid — are known1,2 to be implicated in these ‘amyloid diseases’. Understanding the early steps of fibril formation is critical, and the conditions, mechanism and kinetics of protein and peptide aggregation are being widely investigated through a variety of in vitro studies. Kinetic aspects of the dispersion of the protein or peptide in solution are thought to influence the fibrillization process by mass-transfer effects. In addition, mixing also leads to shear forces, which can influence fibril growth by perturbing the equilibrium between the isolated and aggregated proteins, causing existing fibrils to fragment and create new nuclei3. Writing in the Journal of the American Chemical Society, David Talaga and co-workers have now highlighted4 an additional factor that can influence the fibrillization of amyloid-forming proteins — the presence of hydrophobic interfaces.

Clitic-doubled left dislocation and focus fronting in L2 Spanish: a case of successful acquisition at the syntax-discourse interface

This experimental study tests the Interface Hypothesis by looking into processes at the syntax– discourse interface, teasing apart acquisition of syntactic, semantic and discourse knowledge. Adopting López’s (2009) pragmatic features [±a(naphor)] and [±c(ontrast)], which in combination account for the constructions of dislocation and fronting, we tested clitic left dislocation and fronted focus in the comprehension of English native speakers learning Spanish. Furthermore, we tested knowledge of an additional semantic property: the relationship between the discourse anaphor and the antecedent in clitic left dislocation (CLLD). This relationship is free: it can be subset, superset, part/whole. Syntactic knowledge of clitics was a condition for inclusion in the main test. Our findings indicate that all learners are sensitive to the semantic constraints. While the near-native speakers display native-like discourse knowledge, the advanced speakers demonstrated some discourse knowledge, and intermediate learners did not display any discourse knowledge. The findings support as well as challenge the Interface Hypothesis.

Gradient competence at the syntax-discourse interface

In this article, we present additional support of Duffield's (2003, 2005) distinction between Underlying Competence and Surface Competence. Duffield argues that a more fine-grained distinction between levels of competence and performance is warranted and necessary. While underlying competence is categorical, surface competence is more probabilistic and gradient, being sensitive to lexical and constructional contingencies, including the contextual appropriateness of a given construction. We examine a subset of results from a study comparing native and learner competence of properties at the syntax-discourse interface. Specifically, we look at the acceptability of Clitic Right Dislocation in native and L2 Spanish, in discourse-appropriate context. We argue that Duffield's distinction is a possible explanation of our results.