Fast and efficient automatic memory management for GPUs using compiler-assisted runtime coherence scheme

Exploiting the performance potential of GPUs requires managing the data transfers to and from them efficiently which is an error-prone and tedious task. In this paper, we develop a software coherence mechanism to fully automate all data transfers between the CPU and GPU without any assistance from the programmer. Our mechanism uses compiler analysis to identify potential stale accesses and uses a runtime to initiate transfers as necessary. This allows us to avoid redundant transfers that are exhibited by all other existing automatic memory management proposals. We integrate our automatic memory manager into the X10 compiler and runtime, and find that it not only results in smaller and simpler programs, but also eliminates redundant memory transfers. Tested on eight programs ported from the Rodinia benchmark suite it achieves (i) a 1.06x speedup over hand-tuned manual memory management, and (ii) a 1.29x speedup over another recently proposed compiler--runtime automatic memory management system. Compared to other existing runtime-only and compiler-only proposals, it also transfers 2.2x to 13.3x less data on average.

Basic studies on the role of components of Bacillus megaterium as flotation biocollectors in sulphide mineral separation

Studies were carried out to assess the utility of the cellular and extracellular constituents of Bacillus megaterium for the flotation of sphalerite and galena minerals. Based on the flotation results on the individual minerals, it was observed that sphalerite was preferentially floated compared to galena. A maximum selectivity index (SI) value of 11.7 was achieved in the presence of the soluble fraction of the thermolysed cells, which was higher than that obtained with the intact cells (SI of 6.5) and the insoluble fraction of the thermolysed cells (SI of 9.6). The results of the various enzymatic treatment tests revealed that extracellular DNA played a vital role in the selective flotation of sphalerite. A noteworthy finding was that the single-stranded DNA (ssDNA) had a higher biocollector capacity vis-A -vis the double-stranded DNA (dsDNA), leading to better flotation efficiency. About 95 % recovery of sphalerite could be achieved from the mineral mixture by the combined addition of the ssDNA with the non-DNA components of the bacterial cells, resulting in a maximum SI of 19.1. Calcium and phosphate components of the nutrient media were found to be essential for better selectivity of separation of sphalerite. The mechanisms of microbe-mineral interaction are discussed.

Precise Slicing in Imperative Programs via Term-Rewriting and Abstract Interpretation

We propose a new approach for producing precise constrained slices of programs in a language such as C. We build upon a previous approach for this problem, which is based on term-rewriting, which primarily targets loop-free fragments and is fully precise in this setting. We incorporate abstract interpretation into term-rewriting, using a given arbitrary abstract lattice, resulting in a novel technique for slicing loops whose precision is linked to the power of the given abstract lattice. We address pointers in a first-class manner, including when they are used within loops to traverse and update recursive data structures. Finally, we illustrate the comparative precision of our slices over those of previous approaches using representative examples.

The interaction of N-trifluoroacetylgalactosamine and its derivatives with winged bean (Psophocarpus tetragonolobus) basic agglutinin reveals differential mechanism of their recognition: a fluorine-19 nuclear magnetic resonance study

Here, we show the binding results of a leguminosae lectin, winged bean basic agglutinin (WBA I) to N-trifluoroacetylgalactosamine (NTFAGalN), methyl-alpha-N-trifluoroacetylgalactosamine (Me alpha NTFAGalN) and methyl-beta-tifluoroacetylgalactosamine (Me beta NTFAGalN) using (19) F NMR spectroscopy. No chemical shift difference between the free and bound states for NTFAGalN and Me beta NTFAGalN, and 0.01-ppm chemical shift change for Me alpha NTFAGalN, demonstrate that the Me alpha NTFAGalN has a sufficiently long residence time on the protein binding site as compared to Me beta NTFAGalN and the free anomers of NTFAGalN. The sugar anomers were found in slow exchange with the binding site of agglutinin. Consequently, we obtained their binding parameters to the protein using line shape analyses. Aforementioned analyses of the activation parameters for the interactions of these saccharides indicate that the binding of alpha and beta anomers of NTFAGalN and Me alpha NTFAGalN is controlled enthalpically, while that of Me beta NTFAGalN is controlled entropically. This asserts the sterically constrained nature of the interaction of the Me beta NTFAGalN with WBA I. These studies thus highlight a significant role of the conformation of the monosaccharide ligands for their recognition by WBA I.

Compiler/Runtime Framework for Dynamic Dataflow Parallelization of Tiled Programs

Task-parallel languages are increasingly popular. Many of them provide expressive mechanisms for intertask synchronization. For example, OpenMP 4.0 will integrate data-driven execution semantics derived from the StarSs research language. Compared to the more restrictive data-parallel and fork-join concurrency models, the advanced features being introduced into task-parallelmodels in turn enable improved scalability through load balancing, memory latency hiding, mitigation of the pressure on memory bandwidth, and, as a side effect, reduced power consumption. In this article, we develop a systematic approach to compile loop nests into concurrent, dynamically constructed graphs of dependent tasks. We propose a simple and effective heuristic that selects the most profitable parallelization idiom for every dependence type and communication pattern. This heuristic enables the extraction of interband parallelism (cross-barrier parallelism) in a number of numerical computations that range from linear algebra to structured grids and image processing. The proposed static analysis and code generation alleviates the burden of a full-blown dependence resolver to track the readiness of tasks at runtime. We evaluate our approach and algorithms in the PPCG compiler, targeting OpenStream, a representative dataflow task-parallel language with explicit intertask dependences and a lightweight runtime. Experimental results demonstrate the effectiveness of the approach.

A further discussion on basic equations for two-phase flow: On collision stress of solid phase

The following points are argued: (i) there are two independent kinds of interaction on interfaces, i.e. the interaction between phases and the collision interaction, and the jump relations on interfaces can accordingly be resolved; (ii) the stress in a particle can also be divided into background stress and collision stress corresponding to the two kinds of interaction on interfaces respectively; (iii) the collision stress, in fact, has no jump on interface, so the averaged value of its derivative is equal to the derivative of its averaged value; (iv) the stress of solid phase in the basic equations for two\|phase flow should include the collision stress, while the stress in the expression of the inter\|phase force contains the background one only. Based on the arguments, the strict method for deriving the equations for two\|phase flow developed by Drew, Ishii et al. is generalized to the dense two\|phase flow, which involves the effect of collision stress.

Políticas sociales y vida cotidiana en Argentina : avances, dificultades y un gran desafío: la inclusión de los jóvenes

Resumen: En este artículo se desarrollan, principalmente, cinco dimensiones de análisis: a) la descripción de las principales políticas sociales llevadas adelante en la última década, poniendo especial énfasis en los programas de transferencia de ingresos; b) los cambios en la estructura social y sus impactos en la vida cotidiana; c) los desafíos de una nueva generación de políticas sociales vinculadas a recuperar la movilidad ascendente a través de la inclusión en el mundo del trabajo; d) la realidad de los jóvenes que no estudian ni trabajan como la realidad social más crítica y la tarea para los próximos años y e) algunas propuestas en torno a la inclusión de los jóvenes poniendo especial énfasis en la creación de una red de tutores. El artículo parte del concepto de que se produjeron mejoras importantes en los últimos años pero que quedan muchas cuestiones pendientes que requieren de nuevas ideas y nuevas políticas sociales.

Basic Mechanical Behaviors And Mechanics Of Shear Banding In Bmgs

In this paper, some basic mechanical behaviors of bulk metallic glasses (BMGs) were discussed. It can be found from the discussions that the mechanical behaviors of BMGs are mainly due to the formation and operation of shear bands in BMGs. Furthermore, the relevant mechanics of shear banding were investigated in the paper. The theoretical analysis of deformation coupling thermal softening and free volume creation softening demonstrates that the free volume creation and thermal softening can jointly promote the formation of shear bands in BMGs, and the observed post mortem. shear band width looks more like that governed by free volume creation. (C) 2007 Elsevier Ltd. All rights reserved.