The Paralax Infrastructure: Automatic Parallelization With a Helping Hand

Speeding up sequential programs on multicores is a challenging problem that is in urgent need of a solution. Automatic parallelization of irregular pointer-intensive codes, exempli?ed by the SPECint codes, is a very hard problem. This paper shows that, with a helping hand, such auto-parallelization is possible and fruitful. This paper makes the following contributions: (i) A compiler framework for extracting pipeline-like parallelism from outer program loops is presented. (ii) Using a light-weight programming model based on annotations, the programmer helps the compiler to ?nd thread-level parallelism. Each of the annotations speci?es only a small piece of semantic information that compiler analysis misses, e.g. stating that a variable is dead at a certain program point. The annotations are designed such that correctness is easily veri?ed. Furthermore, we present a tool for suggesting annotations to the programmer. (iii) The methodology is applied to autoparallelize several SPECint benchmarks. For the benchmark with most parallelism (hmmer), we obtain a scalable 7-fold speedup on an AMD quad-core dual processor. The annotations constitute a parallel programming model that relies extensively on a sequential program representation. Hereby, the complexity of debugging is not increased and it does not obscure the source code. These properties could prove valuable to increase the ef?ciency of parallel programming.

Relationships between lameness and feeding behaviour in cubicle-housed Holstein-Friesian dairy cows

The relationship between lameness and feeding behaviour in dairy cows is not yet fully understood. This study examined the effect of lameness on feeding behaviour at two points during lactation. Forty-five Holstein–Friesian dairy cows (average parity 3.3) were housed in cubicle accommodation after calving and fed a total mixed ration (TMR). At approximately 60 and 120 days post partum, 48 h of information on feeding behaviour (including number of meals eaten, meal duration, meal size and feeding rate) was collected for each animal using feed boxes fitted to a data recording system. At the same time points, locomotion scores were recorded for each cow as a measure of lameness (1.0-sound to 4.5-severely lame). Relationships between feeding behaviour and locomotion score were analysed using Residual Maximum Likelihood (REML) analysis. At both time points, cows with higher locomotion scores ate fewer (P < 0.001), larger meals (P < 0.001) and had a shorter total feeding time (P < 0.001). At day 60 post partum, an increase in locomotion score was associated with a decrease in dry matter intake (P < 0.05), but at day 120 post partum no relationship was found between locomotion score and DMI. No relationship was found at either time point between locomotion score and mean meal duration or rate of feeding. The results of this study suggest that the effect of lameness on feeding behaviour in dairy cows does not remain constant across lactation.

Automatic detection and classification of football players

Color segmentation of images usually requires a manual selection and classification of samples to train the system. This paper presents an automatic system that performs these tasks without the need of a long training, providing a useful tool to detect and identify figures. In real situations, it is necessary to repeat the training process if light conditions change, or if, in the same scenario, the colors of the figures and the background may have changed, being useful a fast training method. A direct application of this method is the detection and identification of football players.

Breast-feeding and childhood-onset type 1 diabetes:A pooled analysis of individual participant data from 43 observational studies

OBJECTIVE: To investigate if there is a reduced risk of type 1 diabetes in children breastfed or exclusively breastfed by performing a pooled analysis with adjustment for recognized confounders.
RESEARCH DESIGN AND METHODS: Relevant studies were identified from literature searches using MEDLINE, Web of Science, and EMBASE. Authors of relevant studies were asked to provide individual participant data or conduct prespecified analyses. Meta-analysis techniques were used to combine odds ratios (ORs) and investigate heterogeneity between studies.
RESULTS: Data were available from 43 studies including 9,874 patients with type 1 diabetes. Overall, there was a reduction in the risk of diabetes after exclusive breast-feeding for >2 weeks (20 studies; OR = 0.75, 95% CI 0.64-0.88), the association after exclusive breast-feeding for >3 months was weaker (30 studies; OR = 0.87, 95% CI 0.75-1.00), and no association was observed after (nonexclusive) breast-feeding for >2 weeks (28 studies; OR = 0.93, 95% CI 0.81-1.07) or >3 months (29 studies; OR = 0.88, 95% CI 0.78-1.00). These associations were all subject to marked heterogeneity (I(2) = 58, 76, 54, and 68%, respectively). In studies with lower risk of bias, the reduced risk after exclusive breast-feeding for >2 weeks remained (12 studies; OR = 0.86, 95% CI 0.75-0.99), and heterogeneity was reduced (I(2) = 0%). Adjustments for potential confounders altered these estimates very little.
CONCLUSIONS: The pooled analysis suggests weak protective associations between exclusive breast-feeding and type 1 diabetes risk. However, these findings are difficult to interpret because of the marked variation in effect and possible biases (particularly recall bias) inherent in the included studies.

Automatic FPGA Synthesis of Memory Intensive C-based Kernels

Realising high performance image and signal processing
applications on modern FPGA presents a challenging implementation problem due to the large data frames streaming through these systems. Specifically, to meet the high bandwidth and data storage demands of these applications, complex hierarchical memory architectures must be manually specified
at the Register Transfer Level (RTL). Automated approaches which convert high-level operation descriptions, for instance in the form of C programs, to an FPGA architecture, are unable to automatically realise such architectures. This paper
presents a solution to this problem. It presents a compiler to automatically derive such memory architectures from a C program. By transforming the input C program to a unique dataflow modelling dialect, known as Valved Dataflow (VDF), a mapping and synthesis approach developed for this dialect can
be exploited to automatically create high performance image and video processing architectures. Memory intensive C kernels for Motion Estimation (CIF Frames at 30 fps), Matrix Multiplication (128x128 @ 500 iter/sec) and Sobel Edge Detection (720p @ 30 fps), which are unrealisable by current state-of-the-art C-based synthesis tools, are automatically derived from a C description of the algorithm.

Automatic Parallelization in the Paralax Compiler

The efficient development of multi-threaded software has, for many years, been an unsolved problem in computer science. Finding a solution to this problem has become urgent with the advent of multi-core processors. Furthermore, the problem has become more complicated because multi-cores are everywhere (desktop, laptop, embedded system). As such, they execute generic programs which exhibit very different characteristics than the scientific applications that have been the focus of parallel computing in the past.
Implicitly parallel programming is an approach to parallel pro- gramming that promises high productivity and efficiency and rules out synchronization errors and race conditions by design. There are two main ingredients to implicitly parallel programming: (i) a con- ventional sequential programming language that is extended with annotations that describe the semantics of the program and (ii) an automatic parallelizing compiler that uses the annotations to in- crease the degree of parallelization.
It is extremely important that the annotations and the automatic parallelizing compiler are designed with the target application do- main in mind. In this paper, we discuss the Paralax approach to im- plicitly parallel programming and we review how the annotations and the compiler design help to successfully parallelize generic programs. We evaluate Paralax on SPECint benchmarks, which are a model for such programs, and demonstrate scalable speedups, up to a factor of 6 on 8 cores.

Automatic decomposition and efficient semi-structured meshing of complex solids

In this paper, a novel approach to automatically sub-divide a complex geometry and apply an efficient mesh is presented. Following the identification and removal of thin-sheet regions from an arbitrary solid using the thick/thin decomposition approach developed by Robinson et al. [1], the technique here employs shape metrics generated using local sizing measures to identify long-slender regions within the thick body. A series of algorithms automatically partition the thick region into a non-manifold assembly of long-slender and complex sub-regions. A structured anisotropic mesh is applied to the thin-sheet and long-slender bodies, and the remaining complex bodies are filled with unstructured isotropic tetrahedra. The resulting semi-structured mesh possesses significantly fewer degrees of freedom than the equivalent unstructured mesh, demonstrating the effectiveness of the approach. The accuracy of the efficient meshes generated for a complex geometry is verified via a study that compares the results of a modal analysis with the results of an equivalent analysis on a dense tetrahedral mesh.

Automatic Insertion of Cohesive Elements for Delamination

Composite damage modelling with cohesive elements has initially been limited to the analysis of interface damage or delamination. However, their use is also being extended to the analysis of inplane tensile failure arising from matrix or fibre fracture. These interface elements are typically placed at locations where failure is likely to occur, which infers a certain a priori knowledge of the crack propagation path(s). In the case of a crack jump for example, the location of the jump is usually not obvious, and the simulation would require the placement of cohesive elements at all element faces. A better option, presented here, is to determine the potential location of cohesive elements and insert them during the analysis. The aim of this work is to enable the determination of the crack path, as part of the solution process. A subroutine has been developed and implemented in the commercial finite element package ABAQUS/Standard[1] in order to automatically insert cohesive elements within a pristine model, on the basis of the analysis of the current stress field. Results for the prediction of delamination are presented in this paper.

Stereoselective disposition of fluoxetine and norfluoxetine during pregnancy and breast-feeding

To compare the disposition of fluoxetine and norfluoxetine enantiomers in the mother, foetus and infant.