This paper examines the current status of professional computing education for undergraduate majors and fundamental computer education for non-majors in China. We describe the evolution of computing education over the past decade and its role in meeting society's requirements for IT talent, and we discuss developing trends for computing education at China's universities. We conclude with a description of the Computer Science major at Peking University, one of the nation's leading research-based computer science departments.

We propose a novel heterogeneous scalable desktop grid system, WaveGrid, which uses a peer-to-peer architecture and can satisfy the needs of applications with fast-turnaround requirements. In WaveGrid, hosts self-organize into a timezone-aware overlay network, which supports straightforward, quick resource discovery. Scheduling methods in WaveGrid take heterogeneity into account in selecting scheduling and migration targets. WaveGrid then rides the wave of available cycles by migrating jobs to hosts located in idle night-time zones around the globe. We evaluate WaveGrid using a heterogeneous host CPU power profile based on empirical data collected from the global computing project BOINC. The simulation results show that WaveGrid perform consistently well with fast turnaround time and low migration overhead. It performs much better than other systems with respect to turnaround, stability and minimal impacts on hosts.

We are concerned with the fundamental problem of event ordering in multiplayer peer-to-peer games. Event ordering, even without faults, requires all-to-all message passing with at least two rounds of communication \cite{keidar01}. Multiplayer games add real-time constraints to this scenario. To meet this challenge, we develop an event scoping mechanism that uses N-Trees for event propagation. Unlike traditional application-layer multicast, N-Trees organize peers by their application-level scope of interest, instead of by their delay-based shortest-path tree. This organization allows peers which are close by in the virtual world to order events without needing to communicate with other peers that are farther away. We show the asymptotic analysis of N-Trees indicates that they will perform well for scalable peer-to-peer event ordering. We also provide an analysis of N-Trees in comparison to other distributed architectures for peer-to-peer games.

Peer-to-peer grids that seek to harvest idle cycles available throughout the Internet are vulnerable to hosts that fraudulently accept computational tasks and then maliciously return arbitrary results. Current strategies employed by popular cooperative computing grids, such as SETI@Home, rely heavily on task replication to check results. However, result verification through replication suffers from two potential shortcomings: (1) susceptibility to collusion in which a group of malicious hosts conspire to return the same bad results and (2) high fixed overhead incurred by running redundant copies of the task. In this paper, we first propose a scheme called Quiz to combat collusion. The basic idea of Quiz is to insert indistinguishable quiz tasks with verifiable results known to the client within a package containing several normal tasks. The client can then accept or reject the normal task results based on the correctness of quiz results. Our second contribution is the promotion of trust-based task scheduling in peer-to-peer grids. By coupling a reputation system with the basic verification schemes - Replication and Quiz - a client can potentially avoid malicious hosts and also reduce the overhead of verification for trusted hosts.

Peer-to-peer computing, the harnessing of idle compute cycles throughout the Internet, offers exciting new research challenges in the converging domains of networking and distributed computing. Our system, Cluster Computing on the Fly, seeks to harvest cycles from ordinary users in an open access, non-institutional environment. We conduct a comprehensive study of generic searching methods in a highly dynamic peer-to-peer environment for locating idle cycles for workpile applications which are heavy consumers of cycles. We compare four scalable search methods: expanding ring, advertisement-based, random walk and rendezvous point. We model a variety of workloads, simple scheduling strategies and stabilities of hosts. Our preliminary results show that under light workloads, rendezvous point performs best, while under heavy workloads, its performance falls below the other techniques. We expected rendezvous point to consistently outperform the other search techniques because of its inherent advantage in gathering knowledge about the idle cycles. However in a peer-to-peer environment, which satisfies requests on-demand, large jobs may dominate, resulting in delays for scheduling smaller jobs.

We are developing a distributed architecture for massively-multiplayer games. In this paper, we focus on designing a low-latency event ordering protocol, called NEO, for this architecture. Previous event ordering protocols prevent several types of cheats at the expense of operating at the latency of the slowest player. We broaden the definition of cheating to include four common protocol level cheats and demonstrate how NEO prevents these cheats. At the same time, NEO has a playout latency independent of network conditions and adapts to network congestion to optimize performance.

Current job scheduling systems for massively parallel machines and Beowulf-class compute clusters support batch scheduling involving two classes of queues: prime time vs. non-prime time. Jobs running in these queue classes must satisfy different criteria with respect to job-size, runtime, or other resource needs. These constraints are designed to delay big jobs to non-prime time in order to provide better quality service during the prime time workday hours. This paper surveys existing prime time/non-prime time scheduling policies and investigates the sensitivity of scheduling performance to changes in the jobsize and runtime limits allowed during prime time vs. non-prime time. Our simulation study, using real workload traces from the NASA NAS IBM SP/2 cluster gives strong evidence for the use of specific prime time limits and sheds light oil the performance trade-offs regarding response times, utilization, short term scheduling algorithm (FCFS vs. EASY backfilling), and success and overflow rates.

The multicast address allocation problem requires Internet domains to allocate unique addresses to multicast applications from a globally-shared space. We develop a theoretical framework for multicast allocation algorithms that is influenced by subcube allocation in hypercube computer systems. Based on this framework we derive complexity results for the address allocation problem and describe several new allocation algorithms that use a hypercube model for address representation.

To support IP multicast, domains must assign a unique multicast address to each application from a limited, globally-shared address space. We examine the performance of several classes of address allocation algorithms within the context of the MASC architecture. This study is the first of its kind to model the generalized multicast address allocation problem and consider non-contiguous allocation algorithms. We find that prefix-based allocation outperforms our non-contiguous algorithm, despite the apparent advantages of non-contiguous allocation. We also verify the benefit of using worst-fit for new allocations.

As communication and I/O traffic increase on the interconnection network of high-performance systems, network contention becomes a critical problem drastically reducing performance. Whereas earlier allocation strategies were either sensitive to communication alone or sensitive to I/O alone, we present a new strategy that is sensitive to both communication and I/O. Our new strategy MC-Elongated, strives to achieve (1) the compactness needed to minimize communication-based contention as well as (2) the balance and orientation relative to I/O nodes needed to minimize I/O-based contention. We tested our new strategy using synthetic workloads and a real workload trace of 6087 jobs captured from a 400 node Intel Paragon. Our results show that with respect to system throughput and average job turnaround time, in environments with varying degree of communication and I/O traffic, MC-Elongated outperforms previous allocation strategies that are in use today. Regarding the tension between communication and I/O, our results show that spatial layout is more critical for I/O intensive jobs at lower utilization levels and more critical for communication-intensive jobs at higher utilization levels; and that in general, the impact of I/O traffic is dominant.

"Parallel I/O" is the support of a single parallel application run on many nodes; application data is distributed among the nodes, and is read or written to a single logical file, itself spread across nodes and disks. Parallel I/O is a mapping problem from the data layout in node memory to the file layout on disks. Since the mapping can be quite complicated and involve significant data movement, optimizing the mapping is critical for performance. We discuss our general model of the problem, describe four Collective Buffering algorithms we designed, and report experiments testing their performance on an Intel Paragon and an IBM SP2 both housed at NASA Ames Research Center. Our experiments show improvements of up to two order of magnitude over standard techniques and the potential to deliver peak performance with minimal hardware support.

Current processor allocation techniques for highly parallel systems are typically restricted to contiguous allocation strategies for which performance suffers significantly due to the inherent problem of fragmentation. As a result, message-passing systems have yet to achieve the high utilization levels exhibited by traditional vector supercomputers. We are investigating processor allocation algorithms which lift the restriction on contiguity of processors in order to address the problem of fragmentation. Three noncontiguous processor allocation strategies-paging allocation, random allocation, and the Multiple Buddy Strategy (MBS)-are proposed and studied in this paper. Simulations compare the performance of the noncontiguous strategies with that of several well-known contiguous algorithms. We show that noncontiguous allocation algorithms perform better overall than the contiguous ones, even when message-passing contention is considered. We also present the results of experiments on an Intel Paragon XP/S-15 with 208 nodes that show noncontiguous allocation is feasible with current technologies.

We address the problem of mapping divide-and-conquer programs to mesh connected multicomputers with wormhole or store-and-forward routing. We propose the binomial tree as an efficient model of parallel divide-and-conquer and present two mappings of the binomial tree to the 2D mesh. Our mappings exploit regularity in the communication structure of the divide-and-conquer computation and are also sensitive to the underlying flow control scheme of the target architecture. We evaluate these mappings using new metrics which are extensions of the classical notions of dilation and contention. We introduce the notion of communication slowdown as a measure of the total communication overhead incurred by a parallel computation. We conclude that significant performance gains can be realized when the mapping is sensitive to the flow control scheme of the target architecture.

Current processor allocation techniques for highly parallel systems have thus far been restricted to contiguous allocation strategies for which performance suffers significantly due to the inherent problem of fragmentation. We are investigating processor allocation algorithms which lift the restriction on contiguity of processors in order to address the problem of fragmentation. Three non-contiguous processor allocation strategies: naive, random and the multiple buddy strategy (MBS) are proposed and studied in this paper. Simulations compare the performance of the non-contiguous strategies with that of several well-known contiguous algorithms. We show that non-contiguous allocation algorithms perform better overall than the contiguous ones, even when message-passing contention is considered. We also present the results of experiments on an Intel Paragon XP/S-15 with 208 nodes that show non-contiguous allocation is feasible with current technologies.>

Investigate the problem of static task assignment in distributed computing systems, i.e. given a set of k communicating tasks to be executed on a distributed system of n processors, to which processor should each task be assigned? The author proposes a family of heuristic algorithms for Stone's classic model of communicating tasks whose goal is the minimization of the total execution and communication costs incurred by an assignment. In addition, she augments this model to include interference costs which reflect the degree of incompatibility between two tasks. Whereas high communication costs serve as a force of attraction between tasks, causing them to be assigned to the same processor, interference costs serve as a force of repulsion between tasks, causing them to be distributed over many processors. The inclusion of interference costs in the model yields assignments with greater concurrency, thus overcoming the tendency of Stone's model to assign all tasks to one or a few processors. Simulation results show that the algorithms perform well and in particular, that the highly efficient Simple Greedy Algorithm performs almost as well as more complex heuristic algorithms.>