Accelerating scientific computations with mixed precision algorithms

On modern architectures, the performance of 32-bit operations is often at least twice as fast as the performance of 64-bit operations. By using a combination of 32-bit and 64-bit floating point arithmetic, the performance of many dense and sparse linear algebra algorithms can be significantly enhanced while maintaining the 64-bit accuracy of the resulting solution. The approach presented here can apply not only to conventional processors but also to other technologies such as Field Programmable Gate Arrays (FPGA), Graphical Processing Units (GPU), and the STI Cell BE processor. Results on modern processor architectures and the STI Cell BE are presented.

Program summary

Program title: ITER-REFCatalogue identifier: AECO_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AECO_v1_0.htmlProgram obtainable from: CPC Program Library, Queen's University, Belfast, N. IrelandLicensing provisions: Standard CPC licence, http://cpc.cs.qub.ac.uk/licence/licence.htmlNo. of lines in distributed program, including test data, etc.: 7211No. of bytes in distributed program, including test data, etc.: 41 862Distribution format: tar.gzProgramming language: FORTRAN 77Computer: desktop, serverOperating system: Unix/LinuxRAM: 512 MbytesClassification: 4.8External routines: BLAS (optional)Nature of problem: On modern architectures, the performance of 32-bit operations is often at least twice as fast as the performance of 64-bit operations. By using a combination of 32-bit and 64-bit floating point arithmetic, the performance of many dense and sparse linear algebra algorithms can be significantly enhanced while maintaining the 64-bit accuracy of the resulting solution.Solution method: Mixed precision algorithms stem from the observation that, in many cases, a single precision solution of a problem can be refined to the point where double precision accuracy is achieved. A common approach to the solution of linear systems, either dense or sparse, is to perform the LU factorization of the coefficient matrix using Gaussian elimination. First, the coefficient matrix A is factored into the product of a lower triangular matrix L and an upper triangular matrix U. Partial row pivoting is in general used to improve numerical stability resulting in a factorization PA=LU, where P is a permutation matrix. The solution for the system is achieved by first solving Ly=Pb (forward substitution) and then solving Ux=y (backward substitution). Due to round-off errors, the computed solution, x, carries a numerical error magnified by the condition number of the coefficient matrix A. In order to improve the computed solution, an iterative process can be applied, which produces a correction to the computed solution at each iteration, which then yields the method that is commonly known as the iterative refinement algorithm. Provided that the system is not too ill-conditioned, the algorithm produces a solution correct to the working precision.Running time: seconds/minutes     

Using Developer Activity Data to Enhance Awareness during Collaborative Software Development

Software development is a global activity unconstrained by the bounds of time and space. A major effect of this increasing scale and distribution is that the shared understanding that developers previously acquired by formal and informal face-to-face meetings is difficult to obtain. This paper proposes a shared awareness model that uses information gathered automatically from developer IDE interactions to make explicit orderings of tasks, artefacts and developers that are relevant to particular work contexts in collaborative, and potentially distributed, software development projects. The research findings suggest that such a model can be used to: identify entities (developers, tasks, artefacts) most associated with a particular work context in a software development project; identify relevance relationships amongst tasks, developers and artefacts e.g. which developers and artefacts are currently most relevant to a task or which developers have contributed to a task over time; and, can be used to identify potential bottlenecks in a project through a ‘social graph’ view. Furthermore, this awareness information is captured and provided as developers work in different locations and at different times.     

Design methodology and fabrication process of a microinductor for the next generation of DC–DC power converters

Magnetic materials used in cored microinductors to supersede ferrite in the 0.5-10 MHz frequency range are investigated in this article. The performance of electrodeposited nickel–iron, cobalt–iron–copper alloys and the commercial alloy Vitrovac 6025 have been assessed through their inclusion within a custom-made solenoid microinductor. Although the present inductor achieves 77% power efficiency at 500 KHz for 24.7 W/cm3 power density, an optimized process predicts a power efficiency of 97% for 30.83 W/cm3 power density. The main issues regarding microinductor design and performance are discussed.     

Pupil detection for head-mounted eye tracking in the wild: an evaluation of the state of the art

Robust and accurate detection of the pupil position is a key building block for head-mounted eye tracking and prerequisite for applications on top, such as gaze-based human–computer interaction or attention analysis. Despite a large body of work, detecting the pupil in images recorded under real-world conditions is challenging given significant variability in the eye appearance (e.g., illumination, reflections, occlusions, etc.), individual differences in eye physiology, as well as other sources of noise, such as contact lenses or make-up. In this paper we review six state-of-the-art pupil detection methods, namely ElSe (Fuhl et al. in Proceedings of the ninth biennial ACM symposium on eye tracking research & applications, ACM. New York, NY, USA, pp 123–130, 2016), ExCuSe (Fuhl et al. in Computer analysis of images and patterns. Springer, New York, pp 39–51, 2015), Pupil Labs (Kassner et al. in Adjunct proceedings of the 2014 ACM international joint conference on pervasive and ubiquitous computing (UbiComp), pp 1151–1160, 2014. doi: 10.1145/2638728.2641695), SET (Javadi et al. in Front Neuroeng 8, 2015), Starburst (Li et al. in Computer vision and pattern recognition-workshops, 2005. IEEE Computer society conference on CVPR workshops. IEEE, pp 79–79, 2005), and ?wirski (?wirski et al. in Proceedings of the symposium on eye tracking research and applications (ETRA). ACM, pp 173–176, 2012. doi: 10.1145/2168556.2168585). We compare their performance on a large-scale data set consisting of 225,569 annotated eye images taken from four publicly available data sets. Our experimental results show that the algorithm ElSe (Fuhl et al. 2016) outperforms other pupil detection methods by a large margin, offering thus robust and accurate pupil positions on challenging everyday eye images.     

Relating the X-band opacity of a tropical tree canopy to sapflow, rain interception and dew formation

During summer and autumn 2007, a 11 GHz microwave radiometer was deployed in an experimental tree plantation in Sardinilla, Panama. The opacity of the tree canopy was derived from incoming brightness temperatures received on the ground. A collocated eddy-covariance flux tower measured water vapor fluxes and meteorological variables above the canopy. In addition, xylem sapflow of trees was measured within the flux tower footprint. We observed considerable diurnal differences between measured canopy opacities and modeled theoretical opacities that were closely linked to xylem sapflow. It is speculated that dielectric changes in the leaves induced by the sapflow are causing the observed diurnal changes. In addition, canopy intercepted rain and dew formation also modulated the diurnal opacity cycle. With an enhanced canopy opacity model accounting for water deposited on the leaves, we quantified the influence of canopy stored water (i.e. intercepted water and dew) on the opacity. A time series of dew formation and rain interception was directly monitored during a period of two weeks. We found that during light rainfall up to 60% of the rain amount is intercepted by the canopy whereas during periods of intense rainfall only 4% were intercepted. On average, 0.17 mm of dew was formed during the night. Dew evaporation contributed 5% to the total water vapor flux measured above the canopy.     

TCP-aware network coding with opportunistic scheduling in wireless mobile ad hoc networks

This paper presents a scheme that employs TCP-aware network coding with opportunistic scheduling to enhance TCP throughput in wireless mobile ad hoc networks. Specifically, it considers a TCP parameter, congestion window size, and wireless channel conditions simultaneously to improve TCP throughput performance. Evaluation of this scheme is carried out by using ns2 simulations in different scenarios. The results show that the proposed scheme gives approximately 35% throughput improvement in a high mobility environment and about 33% throughput increase in no or low mobility environment as compared to traditional network coding with opportunistic scheduling. This paper also proposes a new adaptive-W (i.e., adaptive Waiting time) scheme whose objective is to adaptively control waiting time of overheard packets that are stored in a buffer to achieve tradeoff between throughput and overhead.     

Kolmogorov complexity and combinatorial methods in communication complexity

We introduce a method based on Kolmogorov complexity to prove lower bounds on communication complexity. The intuition behind our technique is close to information theoretic methods.We use Kolmogorov complexity for three different things: first, to give a general lower bound in terms of Kolmogorov mutual information; second, to prove an alternative to Yao’s minmax principle based on Kolmogorov complexity; and finally, to identify hard inputs.We show that our method implies the rectangle and corruption bounds, known to be closely related to the subdistribution bound. We apply our method to the hidden matching problem, a relation introduced to prove an exponential gap between quantum and classical communication. We then show that our method generalizes the VC dimension and shatter coefficient lower bounds. Finally, we compare one-way communication and simultaneous communication in the case of distributional communication complexity and improve the previous known result.     

Prioritizing usability problems: data-driven and judgement-driven severity estimates

Software-ergonomic system analysis often reveals numerous usability problems. Given that system design suffers from limited resources, the prioritization of usability problems seems inevitable. Surprisingly enough, prioritization is not in the focus of scientific interest. Within this paper, approaches to prioritization relying on severity estimates will be presented. Two of the approaches, namely priorities based on data about the impact of a problem (data-driven) and priorities based on judgements of interest group members (judgement-driven) will be further explored. In the data-driven approach total problem-handling time caused by a usability problem is presented as a measure of severity. The major disadvantage of the data-driven approach is its costs. A possible alternative are severity estimates based on judgements by members of involved interest groups. The first of two studies shows how to obtain judgement driven severity estimates and reveals a fundamental lack of correspondence between data-driven and judgement-driven severity estimates. The second study supports the notion that the lack of correspondence may stem from a difference between assumptions of the data-driven approach and the naive judgement model of interest group members in the judgement-driven approach. A hypothetical model for severity estimates by interest group members is presented.     

Independent Hybrid Force/Motion Control of Constrained Six-Degrees-of-Freedom Manipulators

Controlling the dynamics of a constrained manipulator includesposition tracking as well as stabilization of the contact wrench.In this paper we derive a control scheme, that makes it possible to treat position and force control independently.The approach is based on a mass-orthogonal splitting of the space of joint torques, allowing independent actuation and therefore independent specification of control laws. An appropriate definition of the reference wrenchmakes it possible to achieve independent stability of the position and force loop.