Análisis de rendimiento de aplicaciones en sistemas multicore

El rápido crecimiento del los sistemas multicore y los diversos enfoques que estos han tomado, permiten que procesos complejos que antes solo eran posibles de ejecutar en supercomputadores, hoy puedan ser ejecutados en soluciones de bajo coste también denominadas "hardware de comodidad". Dichas soluciones pueden ser implementadas usando los procesadores de mayor demanda en el mercado de consumo masivo (Intel y AMD). Al escalar dichas soluciones a requerimientos de cálculo científico se hace indispensable contar con métodos para medir el rendimiento que los mismos ofrecen y la manera como los mismos se comportan ante diferentes cargas de trabajo. Debido a la gran cantidad de tipos de cargas existentes en el mercado, e incluso dentro de la computación científica, se hace necesario establecer medidas "típicas" que puedan servir como soporte en los procesos de evaluación y adquisición de soluciones, teniendo un alto grado de certeza de funcionamiento. En la presente investigación se propone un enfoque práctico para dicha evaluación y se presentan los resultados de las pruebas ejecutadas sobre equipos de arquitecturas multicore AMD e Intel.

Automatic granularity-aware parallelization of programs with predicates, functions, and constraints

Abstract is not available

Multicore Scheduling of Real-Time Irregular Parallel Algorithms in Linux

Face à estagnação da tecnologia uniprocessador registada na passada década, aos principais fabricantes de microprocessadores encontraram na tecnologia multi-core a resposta `as crescentes necessidades de processamento do mercado. Durante anos, os desenvolvedores de software viram as suas aplicações acompanhar os ganhos de performance conferidos por cada nova geração de processadores sequenciais, mas `a medida que a capacidade de processamento escala em função do número de processadores, a computação sequencial tem de ser decomposta em várias partes concorrentes que possam executar em paralelo, para que possam utilizar as unidades de processamento adicionais e completar mais rapidamente. A programação paralela implica um paradigma completamente distinto da programação sequencial. Ao contrário dos computadores sequenciais tipificados no modelo de Von Neumann, a heterogeneidade de arquiteturas paralelas requer modelos de programação paralela que abstraiam os programadores dos detalhes da arquitectura e simplifiquem o desenvolvimento de aplicações concorrentes. Os modelos de programação paralela mais populares incitam os programadores a identificar instruções concorrentes na sua lógica de programação, e a especificá-las sob a forma de tarefas que possam ser atribuídas a processadores distintos para executarem em simultâneo. Estas tarefas são tipicamente lançadas durante a execução, e atribuídas aos processadores pelo motor de execução subjacente. Como os requisitos de processamento costumam ser variáveis, e não são conhecidos a priori, o mapeamento de tarefas para processadores tem de ser determinado dinamicamente, em resposta a alterações imprevisíveis dos requisitos de execução. `A medida que o volume da computação cresce, torna-se cada vez menos viável garantir as suas restrições temporais em plataformas uniprocessador. Enquanto os sistemas de tempo real se começam a adaptar ao paradigma de computação paralela, há uma crescente aposta em integrar execuções de tempo real com aplicações interativas no mesmo hardware, num mundo em que a tecnologia se torna cada vez mais pequena, leve, ubíqua, e portável. Esta integração requer soluções de escalonamento que simultaneamente garantam os requisitos temporais das tarefas de tempo real e mantenham um nível aceitável de QoS para as restantes execuções. Para tal, torna-se imperativo que as aplicações de tempo real paralelizem, de forma a minimizar os seus tempos de resposta e maximizar a utilização dos recursos de processamento. Isto introduz uma nova dimensão ao problema do escalonamento, que tem de responder de forma correcta a novos requisitos de execução imprevisíveis e rapidamente conjeturar o mapeamento de tarefas que melhor beneficie os critérios de performance do sistema. A técnica de escalonamento baseado em servidores permite reservar uma fração da capacidade de processamento para a execução de tarefas de tempo real, e assegurar que os efeitos de latência na sua execução não afectam as reservas estipuladas para outras execuções. No caso de tarefas escalonadas pelo tempo de execução máximo, ou tarefas com tempos de execução variáveis, torna-se provável que a largura de banda estipulada não seja consumida por completo. Para melhorar a utilização do sistema, os algoritmos de partilha de largura de banda (capacity-sharing) doam a capacidade não utilizada para a execução de outras tarefas, mantendo as garantias de isolamento entre servidores. Com eficiência comprovada em termos de espaço, tempo, e comunicação, o mecanismo de work-stealing tem vindo a ganhar popularidade como metodologia para o escalonamento de tarefas com paralelismo dinâmico e irregular. O algoritmo p-CSWS combina escalonamento baseado em servidores com capacity-sharing e work-stealing para cobrir as necessidades de escalonamento dos sistemas abertos de tempo real. Enquanto o escalonamento em servidores permite partilhar os recursos de processamento sem interferências a nível dos atrasos, uma nova política de work-stealing que opera sobre o mecanismo de capacity-sharing aplica uma exploração de paralelismo que melhora os tempos de resposta das aplicações e melhora a utilização do sistema. Esta tese propõe uma implementação do algoritmo p-CSWS para o Linux. Em concordância com a estrutura modular do escalonador do Linux, ´e definida uma nova classe de escalonamento que visa avaliar a aplicabilidade da heurística p-CSWS em circunstâncias reais. Ultrapassados os obstáculos intrínsecos `a programação da kernel do Linux, os extensos testes experimentais provam que o p-CSWS ´e mais do que um conceito teórico atrativo, e que a exploração heurística de paralelismo proposta pelo algoritmo beneficia os tempos de resposta das aplicações de tempo real, bem como a performance e eficiência da plataforma multiprocessador.

Factores de rendimiento en entornos multicore

Este documento refleja el estudio de investigación para la detección de factores que afectan al rendimiento en entornos multicore. Debido a la gran diversidad de arquitecturas multicore se ha definido un marco de trabajo, que consiste en la adopción de una arquitectura específica, un modelo de programación basado en paralelismo de datos, y aplicaciones del tipo Single Program Multiple Data. Una vez definido el marco de trabajo, se han evaluado los factores de rendimiento con especial atención al modelo de programación. Por este motivo, se ha analizado la librería de threads y la API OpenMP para detectar aquellas funciones sensibles de ser sintonizadas al permitir un comportamiento adaptativo de la aplicación al entorno, y que dependiendo de su adecuada utilización han de mejorar el rendimiento de la aplicación.

Multicore-enabling the MPJ express messaging library

With the transition to multicore processors almost complete, the parallel processing community is seeking efficient ways to port legacy message passing applications on shared memory and multicore processors. MPJ Express is our reference implementation of Message Passing Interface (MPI)-like bindings for the Java language. Starting with the current release, the MPJ Express software can be configured in two modes: the multicore and the cluster mode. In the multicore mode, parallel Java applications execute on shared memory or multicore processors. In the cluster mode, Java applications parallelized using MPJ Express can be executed on distributed memory platforms like compute clusters and clouds. The multicore device has been implemented using Java threads in order to satisfy two main design goals of portability and performance. We also discuss the challenges of integrating the multicore device in the MPJ Express software. This turned out to be a challenging task because the parallel application executes in a single JVM in the multicore mode. On the contrary in the cluster mode, the parallel user application executes in multiple JVMs. Due to these inherent architectural differences between the two modes, the MPJ Express runtime is modified to ensure correct semantics of the parallel program. Towards the end, we compare performance of MPJ Express (multicore mode) with other C and Java message passing libraries---including mpiJava, MPJ/Ibis, MPICH2, MPJ Express (cluster mode)---on shared memory and multicore processors. We found out that MPJ Express performs signicantly better in the multicore mode than in the cluster mode. Not only this but the MPJ Express software also performs better in comparison to other Java messaging libraries including mpiJava and MPJ/Ibis when used in the multicore mode on shared memory or multicore processors. We also demonstrate effectiveness of the MPJ Express multicore device in Gadget-2, which is a massively parallel astrophysics N-body siimulation code.

Implementação paralela escalável e eficiente do algoritmo simplex padrão em arquitetura multicore

This work presents a scalable and efficient parallel implementation of the Standard Simplex algorithm in the multicore architecture to solve large scale linear programming problems. We present a general scheme explaining how each step of the standard Simplex algorithm was parallelized, indicating some important points of the parallel implementation. Performance analysis were conducted by comparing the sequential time using the Simplex tableau and the Simplex of the CPLEXR IBM. The experiments were executed on a shared memory machine with 24 cores. The scalability analysis was performed with problems of different dimensions, finding evidence that our parallel standard Simplex algorithm has a better parallel efficiency for problems with more variables than constraints. In comparison with CPLEXR , the proposed parallel algorithm achieved a efficiency of up to 16 times better

Applying the reactive programming paradigm: Toward a more declarative application development approach

After almost 10 years from “The Free Lunch Is Over” article, where the need to parallelize programs started to be a real and mainstream issue, a lot of stuffs did happened: • Processor manufacturers are reaching the physical limits with most of their approaches to boosting CPU performance, and are instead turning to hyperthreading and multicore architectures; • Applications are increasingly need to support concurrency; • Programming languages and systems are increasingly forced to deal well with concurrency. This thesis is an attempt to propose an overview of a paradigm that aims to properly abstract the problem of propagating data changes: Reactive Programming (RP). This paradigm proposes an asynchronous non-blocking approach to concurrency and computations, abstracting from the low-level concurrency mechanisms.

Lock-free Parallel Dynamic Programming

We show a method for parallelizing top down dynamic programs in a straightforward way by a careful choice of a lock-free shared hash table implementation and randomization of the order in which the dynamic program computes its subproblems. This generic approach is applied to dynamic programs for knapsack, shortest paths, and RNA structure alignment, as well as to a state-of-the-art solution for minimizing the máximum number of open stacks. Experimental results are provided on three different modern multicore architectures which show that this parallelization is effective and reasonably scalable. In particular, we obtain over 10 times speedup for 32 threads on the open stacks problem.

Minimizing total tardiness in a stochastic single machine scheduling problem using approximate dynamic programming

This paper addresses the non-preemptive single machine scheduling problem to minimize total tardiness. We are interested in the online version of this problem, where orders arrive at the system at random times. Jobs have to be scheduled without knowledge of what jobs will come afterwards. The processing times and the due dates become known when the order is placed. The order release date occurs only at the beginning of periodic intervals. A customized approximate dynamic programming method is introduced for this problem. The authors also present numerical experiments that assess the reliability of the new approach and show that it performs better than a myopic policy.

ECONOMICAL OCCUPATION OF AGRICULTURAL AREA OF PIRACICABA (SAO PAULO STATE, BRAZIL), INCLUDING RISK THROUGH LINEAR PROGRAMMING

The economic occupation of an area of 500 ha for Piracicaba was studied with the irrigated cultures of maize, tomato, sugarcane and beans, having used models of deterministic linear programming and linear programming including risk for the Target-Motad model, where two situations had been analyzed. In the deterministic model the area was the restrictive factor and the water was not restrictive for none of the tested situations. For the first situation the gotten maximum income was of R$ 1,883,372.87 and for the second situation it was of R$ 1,821,772.40. In the model including risk a producer that accepts risk can in the first situation get the maximum income of R$ 1,883,372. 87 with a minimum risk of R$ 350 year(-1), and in the second situation R$ 1,821,772.40 with a minimum risk of R$ 40 year(-1). Already a producer averse to the risk can get in the first situation a maximum income of R$ 1,775,974.81 with null risk and for the second situation R$ 1.707.706, 26 with null risk, both without water restriction. These results stand out the importance of the inclusion of the risk in supplying alternative occupations to the producer, allowing to a producer taking of decision considered the risk aversion and the pretension of income.

Rough Notes on Programming AVR Microcontrollers in C

These notes follow on from the material that you studied in CSSE1000 Introduction to Computer Systems. There you studied details of logic gates, binary numbers and instruction set architectures using the Atmel AVR microcontroller family as an example. In your present course (METR2800 Team Project I), you need to get on to designing and building an application which will include such a microcontroller. These notes focus on programming an AVR microcontroller in C and provide a number of example programs to illustrate the use of some of the AVR peripheral devices.

Response (re-)programming in aging: A kinematic analysis

Background. Age-related motor slowing may reflect either motor programming deficits, poorer movement execution, or mere strategic preferences for online guidance of movement. We controlled such preferences, limiting the extent to which movements could be programmed. Methods. Twenty-four young and 24 older adults performed a line drawing task that allowed movements to he prepared in advance in one case (i.e., cue initially available indicating target location) and not in another (i.e., no cue initially available as to target location). Participants connected large or small targets illuminated by light-emitting diodes upon a graphics tablet that sampled pen tip position at 200 Hz. Results. Older adults had a disproportionate difficulty initiating movement when prevented from programming in advance. Older adults produced slower, less efficient movements, particularly when prevented from programming under greater precision requirements. Conclusions. The slower movements of older adults do not simply reflect a preference for online control, as older adults have less efficient movements when forced to reprogram their movements. Age-related motor slowing kinematically resembles that seen in patients with cerebellar dysfunction.

Programming internet based DAI applications in Qu-Prolog

This paper presents the unique collection of additional features of Qu-Prolog, a variant of the Al programming language Prolog, and illustrates how they can be used for implementing DAI applications. By this we mean applications comprising communicating information servers, expert systems, or agents, with sophisticated reasoning capabilities and internal concurrency. Such an application exploits the key features of Qu-Prolog: support for the programming of sound non-clausal inference systems, multi-threading, and high level inter-thread message communication between Qu-Prolog query threads anywhere on the internet. The inter-thread communication uses email style symbolic names for threads, allowing easy construction of distributed applications using public names for threads. How threads react to received messages is specified by a disjunction of reaction rules which the thread periodically executes. A communications API allows smooth integration of components written in C, which to Qu-Prolog, look like remote query threads.