Interplay of software bloat, hardware energy proportionality and system bottlenecks

In large flexible software systems, bloat occurs in many forms, causing excess resource utilization and resource bottlenecks. This results in lost throughput and wasted joules. However, mitigating bloat is not easy; efforts are best applied where savings would be substantial. To aid this we develop an analytical model establishing the relation between bottleneck in resources, bloat, performance and power. Analyses with the model places into perspective results from the first experimental study of the power-performance implications of bloat. In the experiments we find that while bloat reduction can provide as much as 40% energy savings, the degree of impact depends on hardware and software characteristics. We confirm predictions from our model with selected results from our experimental study. Our findings show that a software-only view is inadequate when assessing the effects of bloat. The impact of bloat on physical resource usage and power should be understood for a full systems perspective to properly deploy bloat reduction solutions and reap their power-performance benefits.

Reuse, recycle to de-bloat software

Most Java programmers would agree that Java is a language that promotes a philosophy of “create and go forth”. By design, temporary objects are meant to be created on the heap, possibly used and then abandoned to be collected by the garbage collector. Excessive generation of temporary objects is termed “object churn” and is a form of software bloat that often leads to performance and memory problems. To mitigate this problem, many compiler optimizations aim at identifying objects that may be allocated on the stack. However, most such optimizations miss large opportunities for memory reuse when dealing with objects inside loops or when dealing with container objects. In this paper, we describe a novel algorithm that detects bloat caused by the creation of temporary container and String objects within a loop. Our analysis determines which objects created within a loop can be reused. Then we describe a source-to-source transformation that efficiently reuses such objects. Empirical evaluation indicates that our solution can reduce upto 40% of temporary object allocations in large programs, resulting in a performance improvement that can be as high as a 20% reduction in the run time, specifically when a program has a high churn rate or when the program is memory intensive and needs to run the GC often.

On the dynamic contact angle in simulation of impinging droplets with sharp interface methods

Effects of dynamic contact angle models on the flow dynamics of an impinging droplet in sharp interface simulations are presented in this article. In the considered finite element scheme, the free surface is tracked using the arbitrary Lagrangian-Eulerian approach. The contact angle is incorporated into the model by replacing the curvature with the Laplace-Beltrami operator and integration by parts. Further, the Navier-slip with friction boundary condition is used to avoid stress singularities at the contact line. Our study demonstrates that the contact angle models have almost no influence on the flow dynamics of the non-wetting droplets. In computations of the wetting and partially wetting droplets, different contact angle models induce different flow dynamics, especially during recoiling. It is shown that a large value for the slip number has to be used in computations of the wetting and partially wetting droplets in order to reduce the effects of the contact angle models. Among all models, the equilibrium model is simple and easy to implement. Further, the equilibrium model also incorporates the contact angle hysteresis. Thus, the equilibrium contact angle model is preferred in sharp interface numerical schemes.

Foot-mounted INS for everybody - an open-source embedded implementation

We present an open-source, realtime, embedded implementation of a foot-mounted, zero-velocity-update-aided inertial navigation system. The implementation includes both hardware design and software, uses off-the-shelf components and assembly methods, and features a standard USB interface. The software is written in C and can easily be modified to run user implemented algorithms. The hardware design and the software are released under permissive open-source licenses and production files, source code, documentation, and further resources are available at www.openshoe.org. The reproduction cost for a single unit is below $800, with the inertial measurement unit making up the bulk ($700). The form factor of the implementation is small enough for it to be integrated in the sole of a shoe. A performance evaluation of the system shows a position errors for short trajectories (<;100 [m]) of ± 0.2-1% of the traveled distance, depending on the shape of trajectory.

USHA: unified software and hardware architecture for video decoding

Video decoders used in emerging applications need to be flexible to handle a large variety of video formats and deliver scalable performance to handle wide variations in workloads. In this paper we propose a unified software and hardware architecture for video decoding to achieve scalable performance with flexibility. The light weight processor tiles and the reconfigurable hardware tiles in our architecture enable software and hardware implementations to co-exist, while a programmable interconnect enables dynamic interconnection of the tiles. Our process network oriented compilation flow achieves realization agnostic application partitioning and enables seamless migration across uniprocessor, multi-processor, semi hardware and full hardware implementations of a video decoder. An application quality of service aware scheduler monitors and controls the operation of the entire system. We prove the concept through a prototype of the architecture on an off-the-shelf FPGA. The FPGA prototype shows a scaling in performance from QCIF to 1080p resolutions in four discrete steps. We also demonstrate that the reconfiguration time is short enough to allow migration from one configuration to the other without any frame loss.

Interface capturing simulations of acoustically driven bubble dynamics in and near tissue

Tissue injury during therapeutic ultrasound or lithotripsy is thought, in cases, to be due to the action of cavitation bubbles. Assessing this and mitigating it is challenging since bubble dynamics in the complex confinement of tissues or in small blood vessels are challenging to predict. Simulations tools require specialized algorithms to simultaneously represent strong acoustic waves and shocks, topologically complex liquid‐vapor phase boundaries, and the complex viscoelastic material dynamics of tissue. We discuss advances in a simulation tool for such situations. A single‐mesh Eulerian solver is used to solve the governing equations. Special sharpening terms maintain the liquid‐vapor interface in face of the finite numerical dissipation included in the scheme to accurately capture shocks. A recent enhancement to this formulation has significantly improved this interface capturing procedure, which is demonstrated for simulation of the Rayleigh collapse of a bubble. The solver also transports elastic stresses and can thus be used to assess the effects of elastic properties on bubble dynamics. A shock‐induced bubble collapse adjacent to a model elastic tissue is used to demonstrate this and draw some conclusions regarding the injury suppressing role that tissue elasticity might play.

Microscopic dynamics of nanoparticle monolayers at air-water interface

We present results of surface mechanical and particle tracking measurements of nanoparticles trapped at the air-water interface as a function of their areal density. We monitor both the surface pressure (II) and isothermal compression modulus (epsilon) as well as the dynamics of nanoparticle clusters, using fluorescence confocal microscopy while they are compressed to very high density near the two dimensional close packing density Phi similar to 0.82. We observe non-monotonic variation in both epsilon and the dynamic heterogeneity, characterized by the dynamical susceptibility chi(4) with Phi, in such high density monolayers. We provide insight into the underlying nature of such transitions in close packed high density nanoparticle monolayers in terms of the morphology and flexibility of these soft colloidal particles.. We discuss the significance our results in the context of related studies on two dimensional granular or colloidal systems. (C) 2013 Elsevier Inc. All rights reserved.

Effect of Stress and Interface defects on Photo Luminescence of Si nano-crystals embedded in SiO2

Effect of stress and interface defects on photo luminescence property of a silicon nano-crystal (Si-nc) embedded in amorphous silicon dioxide (a-SiO2) are studied in this paper using a self-consistent quantum-continuum based modeling framework. Si-ncs or quantum dots show photoluminescence at room temperature. Whether its origin is due to Si-nc/a-SiO2 interface defects or quantum confinement of carriers in Si-nc is still an outstanding question. Earlier reports have shown that stresses greater than 12 GPa change the indirect energy band gap structure of bulk Si to a direct energy band gap structure. Such stresses are observed very often in nanostructures and these stresses influence the carrier confinement energy significantly. Hence, it is important to determine the effect of stress in addition to the structure of interface defects on photoluminescence property of Si-nc. In the present work, first a Si-nc embedded in a-SiO2 is constructed using molecular dynamics simulation framework considering the actual conditions they are grown so that the interface and residual stress in the structure evolves naturally during formation. We observe that the structure thus created has an interface of about 1 nm thick consisting of 41.95% of defective states mostly Sin+ (n = 0 to 3) coordination states. Further, both the Si-nc core and the embedding matrix are observed to be under a compressive strain. This residual strain field is applied in an effective mass k.p Hamiltonian formulation to determine the energy states of the carriers. The photo luminescence property computed based on the carrier confinement energy and interface energy states associated with defects will be analysed in details in the paper.

Interface characterization of Co2MnGe/Rh2CuSn Heusler multilayers

To address the amount of disorder and interface diffusion induced by annealing, all-Heusler multilayer structures, consisting of ferromagnetic Co2MnGe and nonmagnetic Rh2CuSn layers of varying thicknesses, have been investigated by means of hard x-ray photoelectron spectroscopy and x-ray magnetic circular dichroism. We find evidence for a 4 angstrom thick magnetically dead layer that, together with the identified interlayer diffusion, are likely reasons for the unexpectedly small magnetoresistance found for current-perpendicular-to-plane giant magnetoresistance devices based on this all-Heusler system. We find that diffusion begins already at comparably low temperatures between 200 and 250 degrees C, where Mn appears to be most prone to diffusion.

Near infrared detectors based on HgSe and HgCdSe quantum dots generated at the liquid-liquid interface

HgSe and Hg0.5Cd0.5Se quantum dos (QDs) are synthesized at room temperature by a novel liquid-liquid interface method and their photodetection properties in the near-IR region are investigated. The photodetection properties of our Te-free systems are found to be comparable to those of the previously reported high performance QD vis-IR detectors including HgTe. The present synthesis indicates the cost-effectiveness of selenium based IR detectors owing to the abundance and lower toxicity of selenium compared to tellurium.

An early in-situ stress signature of the AlN-Si pre-growth interface for successful integration of nitrides with (111) Si

The integration of Metal Organic Chemical Vapor Deposition (MOCVD) grown group III-A nitride device stacks on Si (111) substrates is critically dependent on the quality of the first AlN buffer layer grown. A Si surface that is both oxide-free and smooth is a primary requirement for nucleating such layers. A single parameter, the AlN layer growth stress, is shown to be an early (within 50 nm), clear (<0.5 GPa versus > 1GPa), and fail-safe indicator of the pre-growth surface, and the AlN quality required for successful epitaxy. Grain coalescence model for stress generation is used to correlate growth stress, the AlN-Si interface, and crystal quality. (C) 2013 AIP Publishing LLC.

Mechanical and thermal modeling of In-Cu composites for thermal interface materials applications

In-Cu composite solders have been proposed as an effective thermal interface material. Here, finite element analysis and theoretical treatment of their mechanical and thermal behavior is presented. It was determined that the stresses and the strains were concentrated in the narrow and wider In channels, respectively. Furthermore, it is suggested that an In-Cu composite with disk-shaped Cu inclusions may not only further improve the thermal conductivity but may also reduce the stiffness of In-Cu composites in shear.

Effect of Post-Deposition Annealing on the Al2O3/Si(100) Interface Properties

In the present investigation, Al2O3 thin films were deposited onto Si < 100 > substrates by DC reactive magnetron sputtering. The films were annealed in vacuum for one hour at 623, 823 and 1023 K. The composition of the films was quantitatively estimated using X-ray photoelectron spectroscopy (XPS) and the O/Al ratio was found be in the range 1.19 to 1.43. Grazing incidence X-ray diffraction (GIXRD) results revealed that the annealed films are amorphous in nature. Cross sectional transmission electron microscopy (X-TEM) analysis was carried out to study the microstructure and nature of the Al2O3-Si interface as a function of post-deposition annealing. TEM results revealed the presence of nanocrystalline gamma-Al2O3 in the annealed films and an amorphous interface layer was observed at the Al2O3 Si interface. The thickness of the amorphous interface layer was determined from the TEM analysis and the results are discussed.

Extended liaison as an interface between product and process model in assembly

This paper describes the use of liaison to better integrate product model and assembly process model so as to enable sharing of design and assembly process information in a common integrated form and reason about them. Liaison can be viewed as a set, usually a pair, of features in proximity with which process information can be associated. A liaison is defined as a set of geometric entities on the parts being assembled and relations between these geometric entities. Liaisons have been defined for riveting, welding, bolt fastening, screw fastening, adhesive bonding (gluing) and blind fastening processes. The liaison captures process specific information through attributes associated with it. The attributes are associated with process details at varying levels of abstraction. A data structure for liaison has been developed to cluster the attributes of the liaison based on the level of abstraction. As information about the liaisons is not explicitly available in either the part model or the assembly model, algorithms have been developed for extracting liaisons from the assembly model. The use of liaison is proposed to enable both the construction of process model as the product model is fleshed out, as well as maintaining integrity of both product and process models as the inevitable changes happen to both design and the manufacturing environment during the product lifecycle. Results from aerospace and automotive domains have been provided to illustrate and validate the use of liaisons. (C) 2014 Elsevier Ltd. All rights reserved.