Faculty of Computer Science (Fredericton)
Pages
-
-
Visualizing the performance of scientific applications executing with parareal combined space-time parallel approach
-
by Mengru Wang, Recently, manycore and multicore based massively parallel processing architectures, led by GPUs and multicore CPUs, have become the powerful and cost-effective hardware trend in the high performance computing landscape. To incorporate this processing power, various highly efficient parallel algorithms to exploit multilevel parallelism are being designed and developed. Among them, a novel combined space-time parallel approach that uses the parareal algorithm for time parallelization, demonstrates the possible multiplied speedup on top of that achieved with spatial parallelization. This approach is implemented in the EXN/Aero parallel Computational Fluid Dynamics (CFD) solution, developed by Envenio inc. and UNB CFD lab. Notably, the parareal performance with respect to the speedup and convergence when applied into the combined approach, is further complicated with the effects from the parallel spatial solver which hinders effective analysis.
As a complementary tool to traditional experimental performance analysis with respect to speedup and convergence, in this thesis we propose a visual analysis platform, called EXN/Vizer, to help better understand and gain valuable insights into the parareal performance in search for potential optimization strategies. EXN/Vizer is designed as a modular and comprehensive GUI application platform based on the NetBeans Platform technology. Within EXN/Vizer, two layered major visualization modules are implemented using JavaFX toolkit along with several external utility library modules. The bottom layer visualization module is the EXNMonitor Trackboard for dynamically monitoring CFD solver control data including the CFD convergence data and field monitoring point data, which is provided by the CFD application as result datasets. The top layer module is the EXNParareal Dashboard as the general visualization framework for parareal solution process and performance analysis with the parareal profiling datasets. The live tests with EXN/Aero CFD simulations have shown that EXN/Vizer is highly efficient while performing interactive visualization tasks upon heavy dataset load of several thousands of data points, with the runtime heap memory consumption around 100MB at most and 60MB on average. In addition, the simulation solver control trackboard module is constructed for visualizing any SQLite database structured datasets and is also generalizable for other types of database structured datasets., M.C.S. University of New Brunswick, Faculty of Computer Science, 2017.
-
-
WebAssembly in Node.js
-
by Tobias Nießen, Alongside JavaScript, V8 and Node.js have become essential components of contemporary web and cloud applications. With the addition of WebAssembly to the web, developers finally have a fast platform for performance-critical code. However, this addition also introduces new challenges to client and server applications. New application architectures, such as serverless computing, require instantaneous performance without long startup times. This thesis investigates use cases and integration issues of WebAssembly in Node.js, and the performance and quality of WebAssembly compilation in V8 and Node.js. We present the design and implementation of a multi-process shared code cache for Node.js applications, and demonstrate how such a cache can significantly increase application performance, and reduce application startup time, CPU usage, and memory footprint.
-
-
zTrust: adaptive decentralized trust model for QoS selection in electronic marketplaces
-
by Zeinab Noorian, In the absence of legal enforcement procedures for the participants of an
open e-marketplace, trust and reputation systems are central to resisting
threats from malicious agents. Such systems also provide mechanisms for
identifying participants who disseminate unfair ratings. This thesis research
is in the area of modeling trust in multi-agent systems where self-interested
agents intelligently interact to maximize their benefits. The thesis presents
an adaptive intelligent trust model, called zTrust, with multidisciplinary approaches
well-suited for cooperative and competitive electronic commerce in
such a way that individuals could make optimal decisions in selecting business
transaction partners.
zTrust has two essential elements: 1) a two-layered filtering algorithm, the
Prob-Cog model, with an adaptive threshold evaluation procedure, which
combines cognitive and probabilistic views of trust to classify participants
through modelling their subjectivity, behavioural characteristics and environmental
conditions in order to detect unfair advisers, and 2) a two-layered
Trust-Oriented Service Selection (TOSS) framework to assist consumers in discovering providers who maximize their utility by achieving sufficient trust
and fulfilling consumers' preferences on product quality.
We also propose a trust-oriented mechanism built on a game-theoretic basis
well-suited for competitive e-marketplaces where providers might have
limited inventory. The characteristics of such an environment make consumers
concerned with the possibility of losing the opportunity to do business
with good providers whilst providing truthful reputation information about
providers. The proposed mechanism provides consumers with a means to
strategically determine their reporting behaviour by establishing a balance
between the possibility of losing business opportunities because of truthful
reporting and the possibility of not receiving truthful provider information
from advisers if the consumers report untruthfully.
We provide a series of experimental results in a simulated dynamic environment
where agents may be arriving and departing. We illustrate the efficacy
and the robustness of our approaches in important conditions of the cooperative
and competitive marketplaces, as well as in comparison with other
approaches. We observe that zTrust enables consumers to access more accurate
information, even in a hostile environment, and enjoy high profits as
a result of interacting with high-quality providers. Furthermore, our mechanism
promotes honesty amongst providers as dishonest providers would be
isolated and would lose the opportunity to be selected as transaction partners.
Pages
Zircon - This is a contributing Drupal Theme
Design by
WeebPal.
