Algorithm,software, and hardware optimizations for Delaunay mesh generation on simultaneous multithreaded architectures

This article focuses on the optimization of PCDM, a parallel, two-dimensional (2D) Delaunay mesh generation application, and its interaction with parallel architectures based on simultaneous multithreading (SMT) processors. We first present the step-by-step effect of a series of optimizations on performance. These optimizations improve the performance of PCDM by up to a factor of six. They target issues that very often limit the performance of scientific computing codes. We then evaluate the interaction of PCDM with a real SMT-based SMP system, using both high-level metrics, such as execution time, and low-level information from hardware performance counters.     

Elastic Resource Allocation Algorithms for Collaboration Applications

Cloud technologies can provide elasticity to real-time audio and video (A/V) collaboration applications. However, cloud-based collaboration solutions generally operate on a best-effort basis, with neither delivery nor quality guarantees, and high-quality business focused solutions rely on dedicated infrastructure and hardware-based components. This article describes our 2-year of research in the EMD project, which targets to migrate a hardware-based and business focused A/V collaboration solution to a software-based platform hosted in the cloud, providing higher levels of elasticity and reliability. Our focus during this period was an educational collaboration scenario with teachers and students (locally present in the classroom or remotely following the classes). A model of collaboration streaming (e.g. network topology, codecs, stream, streaming workflow, software components) is defined as base for software deployment and preemptive VM allocation techniques. These heuristics are evaluated using a version of the CloudSim simulator extended to generate and simulate realistic collaboration scenarios, to manage network congestion and to monitor a.o. cost and session delay metrics. Our results show that the algorithms reduce costs when compared to previously designed approaches, having an effectiveness of 99% in meeting A/V collaboration setup deadlines, which is a stringent requirement for this collaboration application.     

Cluster formation in a time-varying multi-agent system

We introduce time-varying parameters in a multi-agent clustering model and we derive necessary and sufficient conditions for the occurrence of clustering behavior with respect to a given cluster structure. For periodically varying parameters the clustering conditions may be formulated in a similar way as for the time-invariant model. The results require the individual weights assigned to the agents to be constant. For time-varying weights we illustrate with an example that the obtained results can no longer be applied.     

Taxonomy for the Network and Service Management Research Field

Network and service management has established itself as a research field in the general area of computer networks. However, up to now, no appropriate organization of the field has been carried out in terms of a comprehensive list of terms and topics. In this paper, we introduce a taxonomy for network and service management. With such a taxonomy, it is possible to better understand the landscape of research as well as to reason about possible future challenges and opportunities. As such, in addition to the taxonomy itself, we also present an initial analysis of the field’s past, present, and future, based on the records of papers submitted and accepted in major conferences in the area, as well as a site survey performed through a questionnaire answered by experts from both industry and academia.     

Discovery of an Allosteric Ligand Binding Site in SMYD3 Lysine Methyltransferase

SMYD3 is a multifunctional epigenetic enzyme with lysine methyltransferase activity and various interaction partners. It is implicated in the pathophysiology of cancers but with an unclear mechanism. To discover tool compounds for clarifying its biochemistry and potential as a therapeutic target, a set of drug-like compounds was screened in a biosensor-based competition assay. Diperodon was identified as an allosteric ligand; its R and S enantiomers were isolated, and their affinities to SMYD3 were determined (K_D=42 and 84 μM, respectively). Co-crystallization revealed that both enantiomers bind to a previously unidentified allosteric site in the C-terminal protein binding domain, consistent with its weak inhibitory effect. No competition between diperodon and HSP90 (a known SMYD3 interaction partner) was observed although SMYD3–HSP90 binding was confirmed (K_D=13 μM). Diperodon clearly represents a novel starting point for the design of tool compounds interacting with a druggable allosteric site, suitable for the exploration of noncatalytic SMYD3 functions and therapeutics with new mechanisms of action.     

Degradation behaviors of GaN light-emitting diodes under high-temperature and high-current stressing

A number of commercially available multiple-quantum well (MQW) InGaN/GaN blue LEDs with wavelengths of about 460 nm and a power of 1 mW were stressed at temperatures ranging from 25 to 120 °C at several accelerated DC currents. Both the forward and reverse current voltage characteristics as well as the electroluminescent spectra of the LEDs were monitored. These effects also resulted in the pronounced degradation of light efficiency and device operation lifetime. We found that the degradation of photonic characteristics, correlated very well with the generation-recombination current which is governed by the defect density. The device degradation is much faster at high temperatures. At nominal operation current and at room temperature, the light intensity degradation reaches a saturation level before the light dyes out. These results shed new lights upon the design and lifetime specifications for the emerging commercial solid-state lighting devices.     

Comparative study of the fluid viscosity in tarsal hairy attachment systems of flies and beetles

Wet adhesive systems of insects strongly rely for their function on the formation of capillary bridges with the substrate. Studies on the chemical composition and evaporation dynamics of tarsal secretions strongly suggest a difference in chemistry of secretion in beetles and flies, both possessing hairy attachment devices. This difference is assumed to influence the viscosity of the secretion. Here, we applied a microrheological technique, based on the immersion of nanometric beads in the collected tarsal footprints, to estimate secretion viscosity in a beetle (Coccinella septempunctata) and a fly (Calliphora vicina). Both species studied possess distinct differences in viscosity, the median of which was calculated as 21.8 and 10.9 mPa s, respectively. We further present an approximate theoretical model to calculate the contact formation time of spatula-like terminal contact elements using the viscosity data of the covering fluid. The estimated contact formation time is proportional to the tarsal secretion viscosity and to the square of the contact radius of the contact element.     

10 Short Films About Fog

The exhibit 10 Short Films About Fog, which was part of the Data & Matter Exhibition at Palazzo Bembo during the Venice Architecture Biennale in 2018, documents and expands experiments into the relationship between fog, textiles, and air. The project sought to probe whether the otherwise formless phenomena of fog could be explicitly formed, and how it might assume structures and textures. By revealing the characteristics of fog that fascinate us, yet are latent, the short films also transform the phenomena through magnification, thereby amplifying inherent behaviors and qualities.