P-SOCRATES: A parallel software framework for time-critical many-core systems

Article in Press, Corrected Proof

Semi-Partitioned Scheduling of Fork-Join Tasks using Work-Stealing

13th IEEE/IFIP International Conference on Embedded and Ubiquitous Computing (EUC 2015). 21 to 23, Oct, 2015, Session W1-A: Multiprocessing and Multicore Architectures. Porto, Portugal.

Towards the Combination of Work-Stealing and Semi-Partitioned Scheduling for Parallel Tasks

Paper/Poster presented in Work in Progress Session, 28th GI/ITG International Conference on Architecture of Computing Systems (ARCS 2015). 24 to 26, Mar, 2015. Porto, Portugal.

Scalable and Efficient Configuration of TimeDivision Multiplexed Resources

Consumer-electronics systems are becoming increasingly complex as the number of integrated applications is growing. Some of these applications have real-time requirements, while other non-real-time applications only require good average performance. For cost-efficient design, contemporary platforms feature an increasing number of cores that share resources, such as memories and interconnects. However, resource sharing causes contention that must be resolved by a resource arbiter, such as Time-Division Multiplexing. A key challenge is to configure this arbiter to satisfy the bandwidth and latency requirements of the real-time applications, while maximizing the slack capacity to improve performance of their non-real-time counterparts. As this configuration problem is NP-hard, a sophisticated automated configuration method is required to avoid negatively impacting design time. The main contributions of this article are: 1) An optimal approach that takes an existing integer linear programming (ILP) model addressing the problem and wraps it in a branch-and-price framework to improve scalability. 2) A faster heuristic algorithm that typically provides near-optimal solutions. 3) An experimental evaluation that quantitatively compares the branch-and-price approach to the previously formulated ILP model and the proposed heuristic. 4) A case study of an HD video and graphics processing system that demonstrates the practical applicability of the approach.

Towards the Combination of Work-Stealing and Semi-Partitioned Scheduling for Parallel Tasks

Poster presented in Work in Progress Session, 28th GI/ITG International Conference on Architecture of Computing Systems (ARCS 2015). 24 to 26, Mar, 2015. Porto, Portugal.

Hard real-time multiprocessor scheduling resilient to core failures

Presented at 21st IEEE International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA 2015). 19 to 21, Aug, 2015, pp 122-131. Hong Kong, China.

Non-preemptive and SRP-based fullypreemptive scheduling of real-time Software Transactional Memory

Recent embedded processor architectures containing multiple heterogeneous cores and non-coherent caches renewed attention to the use of Software Transactional Memory (STM) as a building block for developing parallel applications. STM promises to ease concurrent and parallel software development, but relies on the possibility of abort conflicting transactions to maintain data consistency, which in turns affects the execution time of tasks carrying transactions. Because of this fact the timing behaviour of the task set may not be predictable, thus it is crucial to limit the execution time overheads resulting from aborts. In this paper we formalise a FIFO-based algorithm to order the sequence of commits of concurrent transactions. Then, we propose and evaluate two non-preemptive and one SRP-based fully-preemptive scheduling strategies, in order to avoid transaction starvation.

Towards realistic core-failure-resilient scheduling and analysis

Presented at IEEE Real-Time Systems Symposium (RTSS 2015). 1 to 4, Dec, 2015. San Antonio, U.S.A..

Towards realistic core-failure-resilient scheduling and analysis

Presented at IEEE Real-Time Systems Symposium (RTSS 2015). 1 to 4, Dec, 2015. San Antonio, U.S.A..

A multi-objective optimization of the active and reactive resource scheduling at a distribution level in a smart grid context

In the traditional paradigm, the large power plants supply the reactive power required at a transmission level and the capacitors and transformer tap changer were also used at a distribution level. However, in a near future will be necessary to schedule both active and reactive power at a distribution level, due to the high number of resources connected in distribution levels. This paper proposes a new multi-objective methodology to deal with the optimal resource scheduling considering the distributed generation, electric vehicles and capacitor banks for the joint active and reactive power scheduling. The proposed methodology considers the minimization of the cost (economic perspective) of all distributed resources, and the minimization of the voltage magnitude difference (technical perspective) in all buses. The Pareto front is determined and a fuzzy-based mechanism is applied to present the best compromise solution. The proposed methodology has been tested in the 33-bus distribution network. The case study shows the results of three different scenarios for the economic, technical, and multi-objective perspectives, and the results demonstrated the importance of incorporating the reactive scheduling in the distribution network using the multi-objective perspective to obtain the best compromise solution for the economic and technical perspectives.

Dynamic stability metrics for the container loading problem

The Container Loading Problem (CLP) literature has traditionally evaluated the dynamic stability of cargo by applying two metrics to box arrangements: the mean number of boxes supporting the items excluding those placed directly on the floor (M1) and the percentage of boxes with insufficient lateral support (M2). However, these metrics, that aim to be proxies for cargo stability during transportation, fail to translate real-world cargo conditions of dynamic stability. In this paper two new performance indicators are proposed to evaluate the dynamic stability of cargo arrangements: the number of fallen boxes (NFB) and the number of boxes within the Damage Boundary Curve fragility test (NB_DBC). Using 1500 solutions for well-known problem instances found in the literature, these new performance indicators are evaluated using a physics simulation tool (StableCargo), replacing the real-world transportation by a truck with a simulation of the dynamic behaviour of container loading arrangements. Two new dynamic stability metrics that can be integrated within any container loading algorithm are also proposed. The metrics are analytical models of the proposed stability performance indicators, computed by multiple linear regression. Pearson’s r correlation coefficient was used as an evaluation parameter for the performance of the models. The extensive computational results show that the proposed metrics are better proxies for dynamic stability in the CLP than the previous widely used metrics.

Cutting path as a rural postman problem

The Rural Postman Problem (RPP) is a particular Arc Routing Problem (ARP) which consists of determining a minimum cost circuit on a graph so that a given subset of required edges is traversed. The RPP is an NP-hard problem with significant real-life applications. This paper introduces an original approach based on Memetic Algorithms - the MARP algorithm - to solve the RPP and, also deals with an interesting Industrial Application, which focuses on the path optimization for component cutting operations. Memetic Algorithms are a class of Metaheuristics which may be seen as a population strategy that involves cooperation and competition processes between population elements and integrates “social knowledge”, using a local search procedure. The MARP algorithm is tested with different groups of instances and the results are compared with those gathered from other publications. MARP is also used in the context of various real-life applications.

PIASA: A power and interference aware resource management strategy for heterogeneous workloads in cloud data centers

Cloud data centers have been progressively adopted in different scenarios, as reflected in the execution of heterogeneous applications with diverse workloads and diverse quality of service (QoS) requirements. Virtual machine (VM) technology eases resource management in physical servers and helps cloud providers achieve goals such as optimization of energy consumption. However, the performance of an application running inside a VM is not guaranteed due to the interference among co-hosted workloads sharing the same physical resources. Moreover, the different types of co-hosted applications with diverse QoS requirements as well as the dynamic behavior of the cloud makes efficient provisioning of resources even more difficult and a challenging problem in cloud data centers. In this paper, we address the problem of resource allocation within a data center that runs different types of application workloads, particularly CPU- and network-intensive applications. To address these challenges, we propose an interference- and power-aware management mechanism that combines a performance deviation estimator and a scheduling algorithm to guide the resource allocation in virtualized environments. We conduct simulations by injecting synthetic workloads whose characteristics follow the last version of the Google Cloud tracelogs. The results indicate that our performance-enforcing strategy is able to fulfill contracted SLAs of real-world environments while reducing energy costs by as much as 21%.

A framework for scientific computing with GPUs

Dissertação para obtenção do Grau de Mestre em Engenharia Informática