No abstract

SUMMARYThe paper presents a Conservative Time Window (CTW) algorithm tor parallel simulation of discrete event systems. The physical system to be simulated is partitioned into n dmwint sub-systems, each of whkh is represented by an object. The CTW algorithm identifies a time window for each object, such that events occurring in each window are independent of events in other windows and thus they can be processed concurrently. The CTW algorithm was implemented on a shared memory multiprocessor, a Sequent Symmetry S81 with 16 processors. We measured perbrmance of the CTW algorithm on two types of network topologies: feed-brward networks and networks with feedback loops We used t h e metrics to measure perbrmance: speed-up, average number of independent windows detected by the algorithm, and average number of events occurring in each window. We also investigated the impact of various event scheduling policies on performance. The results obtained show that the CTW algorithm produces good performance in many cases.

No abstract

With increasing demands for High Performance Computing (HPC), new ideas and methods are emerged to utilize computing resources more efficiently. Cloud Computing appears to provide benefits such as resource pooling, broad network access and cost efficiency for the HPC applications. However, moving the HPC applications to the cloud can face several key challenges, primarily, the virtualization overhead, multi-tenancy and network latency. Software-Defined Networking (SDN) as an emerging technology appears to pave the road and provide dynamic manipulation of cloud networking such as topology, routing, and bandwidth allocation. This paper presents a new scheme called ASETS which targets dynamic configuration and monitoring of cloud networking using SDN to improve the performance of HPC applications and in particular task scheduling for HPC as a service on the cloud (HPCaaS). Further, SETSA, (SDN-Empowered Task Scheduler Algorithm) is proposed as a novel task scheduling algorithm for the offered ASETS architecture. SETSA monitors the network bandwidth to take advantage of its changes when submitting tasks to the virtual machines. Empirical analysis of the algorithm in different case scenarios show that SETSA has significant potentials to improve the performance of HPCaaS platforms by increasing the bandwidth efficiency and decreasing task turnaround time. In addition, SETSAW, (SETSA Window) is proposed as an improvement of the SETSA algorithm.

Cloud Computing opens a new chapter in Information Technology. It has its roots in internet technology, and like the Internet, it is rapidly and forcefully advancing into a large range of applications and services. While standardization of Cloud Computing is ongoing, there is every indication that cloud technology is here to stay and will cover most sectors of the society, including education. This paper discusses the potentials of CloudBased Education (CBE) in STEM areas to better stimulate and engage students in their pursuit of knowledge and learning. This paper introduces the concept of CloudBased Education for Computer Science (CBECS) and discusses how its framework can be achieved. Further, it shows how the platform can be generalized to use in various STEM areas. The authors argue that the potentials in using Cloud Computing for teaching Computer Science courses are extraordinary since CS has an intimate relationship with the cloud infrastructure. Thus, CBECS can greatly facilitate teaching complex underlying organizations of CS courses such as Operating Systems, Communication Networks, Cyberspace Security, WebBased Applications, Database, and High Performance Computing. While other STEM education can extensively benefit from CBE at the Software as a Service (SaaS) level to present meaningful examples in the lectures, CS courses can move deeper and utilize also the lower levels of cloud services. Such capability can deliver valuable examples and laboratories for CS students to better understand large scale applications and their complexities.