Papers by Claudio de Amorim
A Survey Of Messaging Software Issues And Systems For Myrinet-Based Clusters
Cluster Computing, 2001
Abstract. Several messaging software architectures (MSAs) have been proposed that entirely remove... more Abstract. Several messaging software architectures (MSAs) have been proposed that entirely remove the operating system from the critical communication path, providing direct user-level access to the network interface and avoiding excessive data copying. In this paper we ...
Avaliação de Sistemas de Coerência de Cache em Arquiteturas ODBMs Cliente-Servidor
Page 1. 68 XVII Simpósio Brasileiro de Banco de Dados Avaliaç ao de Sistemas de Coerência de Cac... more Page 1. 68 XVII Simpósio Brasileiro de Banco de Dados Avaliaç ao de Sistemas de Coerência de Cache em Arquiteturas ODBMS Cliente-Servidor Carla Osthoff,Cristiana Bentes,Daniel Ariosto,Marta Mattoso,Claudio L. Amorim ...
Distributed Shared-Memory System - Project Status
ABSTRACT on of the NCP 2 prototype (refer to figure 1) is expected to begin operation during mid-... more ABSTRACT on of the NCP 2 prototype (refer to figure 1) is expected to begin operation during mid-October, 1997. The NCP 2 prototype will consist of a cluster of 8 PowerPC604based workstations extended with our protocol controllers and interconnected via a Myrinet switch [10]. The prototype will run an adapted version of TreadMarks [5] on top of LINUX. PC Myrinet Card PC Myrinet Card PC Myrinet Card PC Myrinet Card PC Myrinet Card PC Myrinet Card PC Myrinet Card Myrinet Card SWITCH MYRINET PC PowerPC604-Based Workstation Figure 1: A 8-node NCP 2 Cluster of workstations extended with Protocol Controllers (PC) 2.1 The Protocol Controller Hardware As shown in figure 2, the NCP 2 protocol controller (PC) is a PCI-compatible board based on Intel 80960 (33 MHz) with 2 Mbytes of DRAM memory, 8 Mbits of memory SRAM (vector memory), bus-snooping module (SNOOPY), and a global clock (1-
Avalia��o de Sistemas de Coer�ncia de Cache em Arquiteturas ODBMs Cliente-Servidor
Sbbd, 2002
Adaptive Techniques for HomeBased Software DSMs
Sbac Pad, 2001
Abstract—This work introduces the HAP system, a software DSM that dynamically adapts to the paral... more Abstract—This work introduces the HAP system, a software DSM that dynamically adapts to the parallel application's sharing patterns. Adaptation is based on a constantly categorization of the sharing experienced by each page. Migratory and producer/consumer pages are managed in a single-writer mode and may be updated, while falsely-shared pages are managed in multiple-writer mode and under invalidated protocol. A home based protocol can potentially benefit from detecting the migratory sharing pattern and ...
In this paper, we introduce the notion of smart transceivers, i.e., transceivers capable of selec... more In this paper, we introduce the notion of smart transceivers, i.e., transceivers capable of selecting convenient transmission ranges that optimize power usage. To this end, we propose and evaluate two novel metrics, namely terrain blockage rate (TB) and useful transmission rate (UT), which can provide important information about the transmission profile of a certain scenario to the transceivers. In particular, TB and UT can be used as parameters of evaluation of protocols and algorithms in terrain with obstacles. Our preliminary experimental results show TB e UT are effective to determine which range the transceivers should use to better optimize the energy consumption while attaining certain degree of connectivity for several scenarios with obstacles. Resumo. Este artigo introduz a noção de transceptores inteligentes, i.e., transceptores capazes de escolher alcances de transmissão que otimizem o uso de energia. Para isso, propomos duas novas métricas denominadas taxa de bloqueio do terreno(TB) e taxa de transmissão útil (UT), que podem fornecer importantes informações sobre o perfil de transmissão do cenário para os transceptores. Em particular, TB e UT podem ser usadas como parâmetros de avaliação de protocolos e de algoritmos em terrenos com obstáculos. Nossos resultados experimentais preliminares mostram que TB e UT são eficazes para se determinar qual alcance que os transceptores devem utilizar para otimizar melhor o consumo de energia e ainda atingir determinado grau de conectividade, para vários cenários com obstáculos .
In this work, we introduce the Cooperative Java Virtual Machine (CoJVM), a new distributed Java r... more In this work, we introduce the Cooperative Java Virtual Machine (CoJVM), a new distributed Java run-time system that enables concurrent Java programs to efficiently execute on clusters of personal computers or workstations. CoJVM implements Java's shared memory model by enabling multiple standard JVMs to work cooperatively and transparently to support a single distributed sharedmemory across the cluster's nodes. Mostly important, CoJVM requires no change to applications written in standard Java. Our experimental results using several Java benchmarks show that CoJVM performance is respectable with speed-ups ranging from 6.1 to 7.8 for a 8-node cluster.
Revisiting the k-Neighbors Connectivity Problem under Practical Indoor Scenarios
A important goal for Topology Control Protocols is to determine the least amount of neighbors nee... more A important goal for Topology Control Protocols is to determine the least amount of neighbors needed to build a fully connected network, aka the k-neighbors connectivity problem. In this work, we estimated the minimum k- neighbors value for a given Wireless Sensor Network (WSN) under practical office scenarios using parameterized simulations that took into account office' size and layout, path
Efficient Support to Scalable Delivery of VoD Systems Using Collocated Access Points
International Conference on Networking, International Conference on Systems and International Conference on Mobile Communications and Learning Technologies (ICNICONSMCL'06), 2006
Concurrency and Computation: Practice and Experience, 2014
Distributed Hash Tables (DHTs) have been used in several applications, but most DHTs have opted t... more Distributed Hash Tables (DHTs) have been used in several applications, but most DHTs have opted to solve lookups with multiple hops, to minimize bandwidth costs while sacrificing lookup latency. This paper presents D1HT, an original DHT which has a peer-to-peer and self-organizing architecture and maximizes lookup performance with reasonable maintenance traffic, and a Quarantine mechanism to reduce overheads caused by volatile peers. We implemented both D1HT and a prominent single-hop DHT, and we performed an extensive and highly representative DHT experimental comparison, followed by complementary analytical studies. In comparison with current single-hop DHTs, our results showed that D1HT consistently had the lowest bandwidth requirements, with typical reductions of up to one order of magnitude, and that D1HT could be used even in popular Internet applications with millions of users. In addition, we ran the first latency experiments comparing DHTs to directory servers, which revealed that D1HT can achieve latencies equivalent to or better than a directory server, and confirmed its greater scalability properties. Overall, our extensive set of results allowed us to conclude that D1HT can provide a very effective solution for a broad range of environments, from large-scale corporate datacenters to widely deployed Internet applications 1,2 .
Lecture Notes in Computer Science, 2001
Hiding communication latency and coherence overhead in software DSMs
ACM SIGOPS Operating Systems Review, 1996
In this paper we propose the use of a PCI-based pro-grammable protocol controller for hiding comm... more In this paper we propose the use of a PCI-based pro-grammable protocol controller for hiding communication and coherence overheads in software DSMs. Our protocol con-troller provides three different types of overhead tolerance: a) moving basic communication and ...
Microprocessing and Microprogramming, 1989
This paper describes the design and realization of MPH, a Hybrid Parallel Machine, which is inten... more This paper describes the design and realization of MPH, a Hybrid Parallel Machine, which is intended to be used as a tool for research in parallel computing. MPH is a multiprocessor system whose processors are arranged according to a hypercubic topology but with interprocessor communication being realized through memory segments which are shared by each pair of processors.
An approach for wireless sensor networks topology control in indoor scenarios
This work presents a new approach for topology control (TC) in wireless sensor networks (WSN) dev... more This work presents a new approach for topology control (TC) in wireless sensor networks (WSN) devised for indoor scenarios with obstacles and no feedback mechanisms. The technique is supported by two novel metrics, namely Blockage Rate and Useful Area Rate, applicable to environments with well-defined obstacles described by Site Specific Propagation models. Simulation results in some realistic scenarios showed that
An approach for wireless sensor networks topology control in indoor scenarios
Proceedings of the 5th International Latin American Networking Conference on - LANC '09, 2009
This work presents a new approach for topology control (TC) in wireless sensor networks (WSN) dev... more This work presents a new approach for topology control (TC) in wireless sensor networks (WSN) devised for indoor scenarios with obstacles and no feedback mechanisms. The technique is supported by two novel metrics, namely Blockage Rate and Useful Area Rate, applicable to environments with well-defined obstacles described by Site Specific Propagation models. Simulation results in some realistic scenarios showed that the technique allows to relate connectivity with transmission power levels and to identify critical transmission power levels. Also, it is shown that the technique is equivalent to an approximated mean value in terms of connectivity aspects.
Lecture Notes in Computer Science, 1997
Carnival is a performance measurement and analysis tool that assists users in understanding the p... more Carnival is a performance measurement and analysis tool that assists users in understanding the performance of DSM applications and protocols. Using traces of program executions, Carnival presents performance data as a hierarchy of execution pro les. During analysis, Carnival automates the inference process that relates performance phenomena to speci c causes in the source code or DSM protocol using techniques that focus on the two most important sources of overhead in DSM systems: waiting time analysis identi es the causes of synchronization overhead, and produces an explanation for each source of waiting time in the program communication analysis identi es the sequence of requests that result in invalidations, and produces an explanation for each source of communication. We describe these techniques and their implementation in TreadMarks, and show h o w t o u s e w aiting time analysis and communication analysis to improve the running time of two programs from the SPLASH application suite when executed on DEC Alphas connected by a DEC Memory Channel network.
Evaluating the DSMIO Cache-Coherence Algorithm in Cluster-Based Parallel ODBMS
Lecture Notes in Computer Science, 2002
In this paper, we assess the performance of DSMIO cachecoherence algorithm implemented in a paral... more In this paper, we assess the performance of DSMIO cachecoherence algorithm implemented in a parallel object-based database management system (ODBMS). The distinguishing feature of DSMIO is its use of the lazy release memory consistency model and ...
D1HT: a distributed one hop hash table
Proceedings 20th IEEE International Parallel & Distributed Processing Symposium, 2006
Page 1. D1HT: A Distributed One Hop Hash Table ∗ Luiz R. Monnerat ,* and Claudio L. Amorim * * C... more Page 1. D1HT: A Distributed One Hop Hash Table ∗ Luiz R. Monnerat ,* and Claudio L. Amorim * * COPPE - Computer and Systems Engineering TI/TI-E&P/STEP Federal University of Rio de Janeiro PETROBRAS {monnerat,amorim}@cos.ufrj.br Abstract ...
2006 18th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD'06), 2006
State-of-the-art software distributed shared-memory systems (SDSMs) provide a cost-effective solu... more State-of-the-art software distributed shared-memory systems (SDSMs) provide a cost-effective solution to run single-program-multiple-data (SPMD) applications on clusters of distributed memory computers.
Conditional execution: an approach for eliminating the basic block barriers
Microprocessing and Microprogramming, 1994
In this paper we present the CONDEX machine model: a VLIW architecture supporting the conditional... more In this paper we present the CONDEX machine model: a VLIW architecture supporting the conditional execution concept. The model consists of multiple and independ.
Uploads
Papers by Claudio de Amorim