Automatic Data Allocation and Buffer Management for Multi-GPU Machines

Multi-GPU machines are being increasingly used in high-performance computing. Each GPU in such a machine has its own memory and does not share the address space either with the host CPU or other GPUs. Hence, applications utilizing multiple GPUs have to manually allocate and manage data on each GPU. Existing works that propose to automate data allocations for GPUs have limitations and inefficiencies in terms of allocation sizes, exploiting reuse, transfer costs, and scalability. We propose a scalable and fully automatic data allocation and buffer management scheme for affine loop nests on multi-GPU machines. We call it the Bounding-Box-based Memory Manager (BBMM). BBMM can perform at runtime, during standard set operations like union, intersection, and difference, finding subset and superset relations on hyperrectangular regions of array data (bounding boxes). It uses these operations along with some compiler assistance to identify, allocate, and manage data required by applications in terms of disjoint bounding boxes. This allows it to (1) allocate exactly or nearly as much data as is required by computations running on each GPU, (2) efficiently track buffer allocations and hence maximize data reuse across tiles and minimize data transfer overhead, and (3) and as a result, maximize utilization of the combined memory on multi-GPU machines. BBMM can work with any choice of parallelizing transformations, computation placement, and scheduling schemes, whether static or dynamic. Experiments run on a four-GPU machine with various scientific programs showed that BBMM reduces data allocations on each GPU by up to 75% compared to current allocation schemes, yields performance of at least 88% of manually written code, and allows excellent weak scaling.

PolyGLoT: A Polyhedral Loop Transformation Framework for a Graphical Dataflow Language

Polyhedral techniques for program transformation are now used in several proprietary and open source compilers. However, most of the research on polyhedral compilation has focused on imperative languages such as C, where the computation is specified in terms of statements with zero or more nested loops and other control structures around them. Graphical dataflow languages, where there is no notion of statements or a schedule specifying their relative execution order, have so far not been studied using a powerful transformation or optimization approach. The execution semantics and referential transparency of dataflow languages impose a different set of challenges. In this paper, we attempt to bridge this gap by presenting techniques that can be used to extract polyhedral representation from dataflow programs and to synthesize them from their equivalent polyhedral representation. We then describe PolyGLoT, a framework for automatic transformation of dataflow programs which we built using our techniques and other popular research tools such as Clan and Pluto. For the purpose of experimental evaluation, we used our tools to compile LabVIEW, one of the most widely used dataflow programming languages. Results show that dataflow programs transformed using our framework are able to outperform those compiled otherwise by up to a factor of seventeen, with a mean speed-up of 2.30x while running on an 8-core Intel system.

Fast Computing For Lurr Of Earthquake Prediction

The LURR theory is a new approach for earthquake prediction, which achieves good results in earthquake prediction within the China mainland and regions in America, Japan and Australia. However, the expansion of the prediction region leads to the refinement of its longitude and latitude, and the increase of the time period. This requires increasingly more computations, and the volume of data reaches the order of GB, which will be very difficult for a single CPU. In this paper, a new method was introduced to solve this problem. Adopting the technology of domain decomposition and parallelizing using MPI, we developed a new parallel tempo-spatial scanning program.

Peces continentales de la Argentina: Iconografía. Gymnocharacinus bergii.

Espanol: Este trabajo tiene como objetivo la recopilación de dibujos y diferentes figuras de las especies presentes en nuestro territorio. Para ello, hemos utilizado como principales fuentes de información la obra de Ringuelet et al. (1967), Reis et al. (2003), López et al. (2003; 2006) y las bases de datos on-line de W. N. Eschmeyer y Fish Base. El tratamiento incluye una ficha individual con imágenes seleccionadas de la especie tratada y las que eventualmente hayan surgido por trabajos de índole anatómica. Se adjuntan las referencias de los trabajos utilizados en la confección de cada ficha. Esta es una publicación abierta, por lo que requerirá de actualización permanente, para lo cual sería necesario se sume a nuestra tarea la buena voluntad y colaboración de la comunidad ictiológica. Los compiladores entendemos que esta contribución, además de aportar valiosa información, rescata el trabajo de mujeres y hombres que son parte de la rica historia de la ictiología nacional y regional. (Texto en Espanol y PDF tiene viente dos paginas) English: The goal of this work is to present a collection of drawings and figures depicting the species present in our country. To achieve this, our main sources of information have been the works of Ringuelet et al. (1967), Reis et al. (2003), and López et al. (2003; 2006), as well as the online databases of W. N. Eschmeyer and Fish Base. Each species has an individual factsheet with selected images, including some from anatomical research works. All the works referenced are mentioned in each factsheet. This is an open-ended publication. As such, it will require permanent updating, which will depend on the good will and collaboration of the ichthyological community. As compilers, we understand that this contribution not only provides valuable information, but also highlights the work of men and women who are part of the rich history of our national and regional ichthyology. (Text is in Spanish. PDF contains 22 pages.)

Peces continentales de la Argentina: Iconografía. Lepidosiren paradoza.

Spanish: Este trabajo tiene como objetivo la recopilación de dibujos y diferentes figuras de las especies presentes en nuestro territorio. Para ello, hemos utilizado como principales fuentes de información la obra de Ringuelet et al. (1967), Reis et al. (2003), López et al. (2003; 2006) y las bases de datos on-line de W. N. Eschmeyer y Fish Base. El tratamiento incluye una ficha individual con imágenes seleccionadas de la especie tratada y las que eventualmente hayan surgido por trabajos de índole anatómica. Se adjuntan las referencias de los trabajos utilizados en la confección de cada ficha. Esta es una publicación abierta, por lo que requerirá de actualización permanente, para lo cual sería necesario se sume a nuestra tarea la buena voluntad y colaboración de la comunidad ictiológica. Los compiladores entendemos que esta contribución, además de aportar valiosa información, rescata el trabajo de mujeres y hombres que son parte de la rica historia de la ictiología nacional y regional. English: The goal of this work is to present a collection of drawings and figures depicting the species present in our country. To achieve this, our main sources of information have been the works of Ringuelet et al. (1967), Reis et al. (2003), and López et al. (2003; 2006), as well as the online databases of W. N. Eschmeyer and Fish Base. Each species has an individual factsheet with selected images, including some from anatomical research works. All the works referenced are mentioned in each factsheet. This is an open-ended publication. As such, it will require permanent updating, which will depend on the good will and collaboration of the ichthyological community. As compilers, we understand that this contribution not only provides valuable information, but also highlights the work of men and women who are part of the rich history of our national and regional ichthyology. (Texto en espanol y PDF tiene cuatorce paginas.)

Proceedings of the Eighteenth International Sea Turtle Symposium, 3-7 March 1998, Mazatlán, Sinaloa, Mexico

For the first time in its history, the International Symposium on Sea Turtle Biology and Conservation migrated to a site outside of the United States. Thus the Eighteenth edition was hosted by the Mazatlán Research Unit of the Instituto de Ciencias del Mar y Limnología of the Mexican National Autonomous University (UNAM) in Mazatlán, Sinaloa (Mexico) where it was held from 3-7, March, 1998. Above all, our symposium is prominent for its dynamism and enthusiasm in bringing together specialists from the world´s sea turtle populations. In an effort to extend this philosophy, and fully aware of how fast the interest in sea turtles has grown, the organizers paid special attention to bring together as many people as possible. With the tremendous efforts of the Travel Committee and coupled with a special interest by the Latin American region´s devotees, we managed to get 653 participants from 43 countries. The number of presentations increased significantly too, reaching a total of 265 papers, ranging from cutting-edge scientific reports based on highly sophisticated methods, to the experiences and successes of community-based and environmental education programs. A priority given by this symposium was the support and encouragement for the construction of "bridges" across cultural and discipline barriers. We found success in achieving a multinational dialogue among interest groups- scientists, resource managers, decision makers, ngo's, private industry. There was a broad representation of the broad interests that stretch across these sectors, yet everyone was able to listen and offer their own best contribution towards the central theme of the Symposium: the conservation of sea turtles and the diversity of marine and coastal environments in which they develop through their complicated and protracted life cycle. Our multidisciplinary approach is highly important at the present, finding ourselves at a cross roads of significant initiatives in the international arena of environmental law, where the conservation of sea turtles has a key role to play. Many, many people worked hard over the previous 12 months, to make the symposium a success. Our sincerest thanks to all of them: Program committee: Laura Sarti (chair), Ana Barragán, Rod Mast, Heather Kalb, Jim Spotilla, Richard Reina, Sheryan Epperly, Anna Bass, Steve Morreale, Milani Chaloupka, Robert Van Dam, Lew Ehrhart, J. Nichols, David Godfrey, Larry Herbst, René Márquez, Jack Musick, Peter Dutton, Patricia Huerta, Arturo Juárez, Debora Garcia, Carlos Suárez, German Ramírez, Raquel Briseño, Alberto Abreu; Registration and Secretary: Jane Provancha (chair), Lupita Polanco; Informatics: Germán Ramírez, Carlos Suárez; Cover art: Blas Nayar; Designs: Germán Ramírez, Raquel Briseño, Alberto Abreu. Auction: Rod Mast; Workshops and special meetings: Selina Heppell; Student prizes: Anders Rhodin; Resolutions committee: Juan Carlos Cantú; Local organizing committee: Raquel Briseño, Jane Abreu; Posters: Daniel Ríos and Jeffrey Semminoff; Travel committee: Karen Eckert (chair), Marydele Donnelly, Brendan Godley, Annette Broderick, Jack Frazier; Student travel: Francisco Silva and J. Nichols; Vendors: Tom McFarland and J. Nichols; Volunteer coordination: Richard Byles; Latin American Reunión: Angeles Cruz Morelos; Nominations committee: Randall Arauz, Colleen Coogan, Laura Sarti, Donna Shaver, Frank Paladino. Once again, Ed Drane worked his usual magic with the Treasury of the Symposium Significant financial contributions were generously provided by government agencies. SEMARNAP (Mexico´s Ministry of Environment, Natural Resources and Fisheries) through its central office, the Mazatlán Regional Fisheries Research Center (CRIP-Mazatlán) and the National Center for Education and Capacity Building for Sustainable Development (CECADESU) contributed to the logistics and covered the costs of auditoria and audiovisual equipment for the Symposium, teachers and their hotels for the Community Development and Environmental Education workshop in the 5th Latin American Sea Turtle Specialists; DIF (Dept of Family Affairs) provided free accomodation and food for the more than 100 participants in the Latin American Reunion. In this Reunion, the British Council-Mexico sponsored the workshop on the Project Cycle. The National Chamber of the Fisheries Industry (CANAINPES) kindly sponsored the Symposium´s coffee breaks. Personnel from the local Navy (Octave Zona Naval) provided invaluable aid in transport and logistics. The Scientific Coordination Office from UNAM (CICUNAM) and the Latin American Biology Network (RELAB) also provided funding. Our most sincere recognition to all of them. In the name of this Symposium´s compilers, I would like to also express our gratitude to Wayne Witzell, Technical Editor for his guidance and insights and to Jack Frazier for his help in translating and correcting the English of contributions from some non-native English speakers. Many thanks to Angel Fiscal and Tere Martin who helped with the typing in the last, last corrections and editions for these Proceedings. To all, from around the world, who generously helped make the 18th Symposium a huge success, shared their experiences and listened to ours, our deepest gratitude! (PDF contains 316 pages)

Peces continentales de la Argentina: Iconografía. Brycon orbignyanus.

The goal of this work is to present a collection of drawings and figures depicting the species present in our country. To achieve this, our main sources of information have been the works of Ringuelet et al. (1967), Reis et al. (2003), and López et al. (2003; 2006), as well as the online databases of W. N. Eschmeyer and Fish Base. Each species has an individual factsheet with selected images, including some from anatomical research works. All the works referenced are mentioned in each factsheet. This is an open-ended publication. As such, it will require permanent updating, which will depend on the good will and collaboration of the ichthyological community. As compilers, we understand that this contribution not only provides valuable information, but also highlights the work of men and women who are part of the rich history of our national and regional ichthyology.

Aplicação das fórmulas de Vincenty nos cálculos das correções dos efeitos do relevo na gravidade e na altura geoidal.

O presente trabalho apresenta a aplicação das fórmulas de Vincenty nos cálculos das correções do terreno e do efeito indireto, que desempenham papel relevante na construção de cartas geoidais. Implementa-se um programa de processamento que realiza a integração numérica sobre o modelo digital do terreno, discretizado em células triangulares de Delaunay. O sistema foi desenvolvido com a linguagem de programação FORTRAN, para a execução de intensos algoritmos numéricos usando compiladores livres e robustos. Para o cálculo do efeito indireto, considera-se a redução gravimétrica efetuada com base no segundo método de condensação de Helmert, face ao pequeno valor de efeito indireto no cálculo do geóide, em função da mudança que este produz no potencial da gravidade devido ao deslocamento da massa topográfica. Utiliza-se, o sistema geodésico SIRGAS 2000 como sistema de referência para o cômputo das correções. Simplificando o exame dos resultados alcançados, distingue-se o processamento e desenvolvimento do trabalho em etapas como a escolha de ferramentas geodésicas para máxima precisão dos resultados, elaboração de subrotinas e comparação de resultados com cálculos anteriores. Os resultados encontrados foram de geração sadia e satisfatória e podem ser perfeitamente empregados no cálculo do geóide em qualquer área do globo.

Peces continentales de la Argentina: Iconografía. Cheirodon interruptus

The goal of this work is to present a collection of drawings and figures depicting the species present in our country. To achieve this, our main sources of information have been the works of Ringuelet et al. (1967), Reis et al. (2003), and López et al. (2003; 2006), as well as the online databases of W. N. Eschmeyer and Fish Base. Each species has an individual factsheet with selected images, including some from anatomical research works. All the works referenced are mentioned in each factsheet. This is an open-ended publication. As such, it will require permanent updating,which will depend on the good will and collaboration of the ichthyological community. As compilers, we understand that this contribution not only provides valuable information, but also highlights the work of men and women who are part of the rich history of our national and regional ichthyology.

Historical knowledge of sharks: ancient science, earliest American encounters, and American science, fisheries, and utilization

In western civilization, the knowledge of the elasmobranch or selachian fishes (sharks and rays) begins with Aristotle (384–322 B.C.). Two of his extant works, the “Historia Animalium” and the “Generation of Animals,” both written about 330 B.C., demonstrate knowledge of elasmobranch fishes acquired by observation. Roman writers of works on natural history, such as Aelian and Pliny, who followed Aristotle, were compilers of available information. Their contribution was that they prevented the Greek knowledge from being lost, but they added few original observations. The fall of Rome, around 476 A.D., brought a period of economic regression and political chaos. These in turn brought intellectual thought to a standstill for nearly one thousand years, the period known as the Dark Ages. It would not be until the middle of the sixteenth century, well into the Renaissance, that knowledge of elasmobranchs would advance again. The works of Belon, Salviani, Rondelet, and Steno mark the beginnings of ichthyology, including the study of sharks and rays. The knowledge of sharks and rays increased slowly during and after the Renaissance, and the introduction of the Linnaean System of Nomenclature in 1735 marks the beginning of modern ichthyology. However, the first major work on sharks would not appear until the early nineteenth century. Knowledge acquired about sea animals usually follows their economic importance and exploitation, and this was also true with sharks. The first to learn about sharks in North America were the native fishermen who learned how, when, and where to catch them for food or for their oils. The early naturalists in America studied the land animals and plants; they had little interest in sharks. When faunistic works on fishes started to appear, naturalists just enumerated the species of sharks that they could discern. Throughout the U.S. colonial period, sharks were seldom utilized for food, although their liver oil or skins were often utilized. Throughout the nineteenth century, the Spiny Dogfish, Squalus acanthias, was the only shark species utilized in a large scale on both coasts. It was fished for its liver oil, which was used as a lubricant, and for lighting and tanning, and for its skin which was used as an abrasive. During the early part of the twentieth century, the Ocean Leather Company was started to process sea animals (primarily sharks) into leather, oil, fertilizer, fins, etc. The Ocean Leather Company enjoyed a monopoly on the shark leather industry for several decades. In 1937, the liver of the Soupfin Shark, Galeorhinus galeus, was found to be a rich source of vitamin A, and because the outbreak of World War II in 1938 interrupted the shipping of vitamin A from European sources, an intensive shark fishery soon developed along the U.S. West Coast. By 1939 the American shark leather fishery had transformed into the shark liver oil fishery of the early 1940’s, encompassing both coasts. By the late 1940’s, these fisheries were depleted because of overfishing and fishing in the nursery areas. Synthetic vitamin A appeared on the market in 1950, causing the fishery to be discontinued. During World War II, shark attacks on the survivors of sunken ships and downed aviators engendered the search for a shark repellent. This led to research aimed at understanding shark behavior and the sensory biology of sharks. From the late 1950’s to the 1980’s, funding from the Office of Naval Research was responsible for most of what was learned about the sensory biology of sharks.

LAPACK的自动并行化工具研究

*

automated test program generation for an industrial optimizing compiler

Association for Computing Machinery, ACM; IEEE; IEEE Computer Society; SIGSOFT

RABBIT: A Compiler for SCHEME

We have developed a compiler for the lexically-scoped dialect of LISP known as SCHEME. The compiler knows relatively little about specific data manipulation primitives such as arithmetic operators, but concentrates on general issues of environment and control. Rather than having specialized knowledge about a large variety of control and environment constructs, the compiler handles only a small basis set which reflects the semantics of lambda-calculus. All of the traditional imperative constructs, such as sequencing, assignment, looping, GOTO, as well as many standard LISP constructs such as AND, OR, and COND, are expressed in macros in terms of the applicative basis set. A small number of optimization techniques, coupled with the treatment of function calls as GOTO statements, serve to produce code as good as that produced by more traditional compilers. The macro approach enables speedy implementation of new constructs as desired without sacrificing efficiency in the generated code. A fair amount of analysis is devoted to determining whether environments may be stack-allocated or must be heap-allocated. Heap-allocated environments are necessary in general because SCHEME (unlike Algol 60 and Algol 68, for example) allows procedures with free lexically scoped variables to be returned as the values of other procedures; the Algol stack-allocation environment strategy does not suffice. The methods used here indicate that a heap-allocating generalization of the "display" technique leads to an efficient implementation of such "upward funargs". Moreover, compile-time optimization and analysis can eliminate many "funargs" entirely, and so far fewer environment structures need be allocated at run time than might be expected. A subset of SCHEME (rather than triples, for example) serves as the representation intermediate between the optimized SCHEME code and the final output code; code is expressed in this subset in the so-called continuation-passing style. As a subset of SCHEME, it enjoys the same theoretical properties; one could even apply the same optimizer used on the input code to the intermediate code. However, the subset is so chosen that all temporary quantities are made manifest as variables, and no control stack is needed to evaluate it. As a result, this apparently applicative representation admits an imperative interpretation which permits easy transcription to final imperative machine code. These qualities suggest that an applicative language like SCHEME is a better candidate for an UNCOL than the more imperative candidates proposed to date.

Structure Sharing and Parallelization in a GB Parser

By utilizing structure sharing among its parse trees, a GB parser can increase its efficiency dramatically. Using a GB parser which has as its phrase structure recovery component an implementation of Tomita's algorithm (as described in [Tom86]), we investigate how a GB parser can preserve the structure sharing output by Tomita's algorithm. In this report, we discuss the implications of using Tomita's algorithm in GB parsing, and we give some details of the structuresharing parser currently under construction. We also discuss a method of parallelizing a GB parser, and relate it to the existing literature on parallel GB parsing. Our approach to preserving sharing within a shared-packed forest is applicable not only to GB parsing, but anytime we want to preserve structure sharing in a parse forest in the presence of features.

Wireless and Physical Security via Embedded Sensor Networks

Wireless Intrusion Detection Systems (WIDS) monitor 802.11 wireless frames (Layer-2) in an attempt to detect misuse. What distinguishes a WIDS from a traditional Network IDS is the ability to utilize the broadcast nature of the medium to reconstruct the physical location of the offending party, as opposed to its possibly spoofed (MAC addresses) identity in cyber space. Traditional Wireless Network Security Systems are still heavily anchored in the digital plane of "cyber space" and hence cannot be used reliably or effectively to derive the physical identity of an intruder in order to prevent further malicious wireless broadcasts, for example by escorting an intruder off the premises based on physical evidence. In this paper, we argue that Embedded Sensor Networks could be used effectively to bridge the gap between digital and physical security planes, and thus could be leveraged to provide reciprocal benefit to surveillance and security tasks on both planes. Toward that end, we present our recent experience integrating wireless networking security services into the SNBENCH (Sensor Network workBench). The SNBENCH provides an extensible framework that enables the rapid development and automated deployment of Sensor Network applications on a shared, embedded sensing and actuation infrastructure. The SNBENCH's extensible architecture allows an engineer to quickly integrate new sensing and response capabilities into the SNBENCH framework, while high-level languages and compilers allow novice SN programmers to compose SN service logic, unaware of the lower-level implementation details of tools on which their services rely. In this paper we convey the simplicity of the service composition through concrete examples that illustrate the power and potential of Wireless Security Services that span both the physical and digital plane.