Cellular automata (CA) are simulation models that can produce complex emergent behaviors from simple local rules. Although state-of-the-art GPU solutions are already fast due to their data-parallel nature, their performance can rapidly degrade in CA with a large neighborhood radius. With the inclusion of tensor cores across the entire GPU ecosystem, interest has grown in finding ways to leverage these fast units outside the field of artificial intelligence, which was their original purpose. In this work, we present CAT, a GPU tensor core approach that can accelerate CA in which the cell transition function acts on a weighted summation of its neighborhood. CAT is evaluated theoretically, using an extended PRAM cost model, as well as empirically using the Larger Than Life (LTL) family of CA as case studies. The results confirm that the cost model is accurate, showing that CAT exhibits constant time throughout the entire radius range$1 \leq r \leq 16$, and its theoretical speedups agree with the empirical results. At low radius$r=1,2$, CAT is competitive and is only surpassed by the fastest state-of-the-art GPU solution. Starting from$r=3$, CAT progressively outperforms all other approaches, reaching speedups of up to$101\times$over a GPU baseline and up to$\sim \!14\times$over the fastest state-of-the-art GPU approach. In terms of energy efficiency, CAT is competitive in the range$1 \leq r \leq 4$and from$r \geq 5$it is the most energy efficient approach. As for performance scaling across GPU architectures, CAT shows a promising trend that, if continues for future generations, it would increase its performance at a higher rate than classical GPU solutions. A CPU version of CAT was also explored, using the recently introduced AMX instructions. Although its performance is still below GPU tensor cores, it is a promising approach as it can still outperform some GPU approaches at large radius. The results obtained in this work put CAT as an approach with great potential for scientists who need to study emerging phenomena in CA with a large neighborhood radius, both in the GPU and in the CPU.

Persistence diagrams (PD)s play a central role in topological data analysis, and are used in an ever increasing variety of applications. The comparison of PD data requires computing distances among large sets of PDs, with metrics which are accurate, theoretically sound, and fast to compute. Especially for denser multi-dimensional PDs, such comparison metrics are lacking. While on the one hand, Wasserstein-type distances have high accuracy and theoretical guarantees, they incur high computational cost. On the other hand, distances between vectorizations such as Persistence Statistics (PS)s have lower computational cost, but lack the accuracy guarantees and theoretical properties of a true distance over PD space. In this work we introduce a class of pseudodistances called Extended Topological Pseudodistances (ETD)s, which have tunable complexity, and can approximate Sliced and classical Wasserstein distances at the high-complexity extreme, while being computationally lighter and close to Persistence Statistics at the lower complexity extreme, and thus allow users to interpolate between the two metrics. We build theoretical comparisons to show how to fit our new distances at an intermediate level between persistence vectorizations and Wasserstein distances. We also experimentally verify that ETDs outperform PSs in terms of accuracy and outperform Wasserstein and Sliced Wasserstein distances in terms of computational complexity.

Parents that homeschool their children ignore certain topics when they lack mastery or interest in them. Homeschool groups try to address this issue, cooperatively educating their children. We were contacted by such a group that wanted to introduce their children to computational thinking (CT). These children, aged 7-11, have showed an interest in technology, and use online educational resources. None of the parents felt capable of tutoring the group about CT. They also worried about losing control about how their children interact with technology. We report an intervention over 9 months to introduce eleven young homeschoolers to CT in a suburban environment, describing the impact on parent and children attitudes towards technology and CT. We conclude with three lessons: 1)~science-related activities should be used to introduce CT among homeschoolers, 2) "success'' is establishing a meaningful relationship with a homeschool group, and 3) activities designed for school children need to be adapted to the homeschooling context.

In Latin America, computational thinking workshops are mostly developed in urban areas, charging participation fees. And although teachers are increasingly being expected to include technology in their classrooms, computational thinking and programming are not mandatory topics in teacher training programs. This hinders the development of digital skills among Latino students, and we expect that the digital gap between urban and rural populations will expand over time, especially in socio-economically vulnerable populations. Believing that teachers can be agents of change, we designed a 12 hour workshop to train the K-8 teaching staff in a rural and vulnerable school. The goal of this experience was to help these teachers develop basic computational thinking skills and devise new ways to incorporate what they learned in their classrooms. In this paper, we report our experience facilitating this workshop, and analyze the teacher perceptions before and after the intervention. Teacher attitudes changed drastically during the execution of the workshop, and they were able to come up with creative ways of incorporating computational thinking activities into their subjects. The reported experience can be used as input to develop public policies with respect to how computational thinking should be introduced in rural and vulnerable environments.

A Bilingual Multimodal Speech Communication Corpus incorporating acoustic data as well as visual data related to face, hands and arms gestures during speech, is presented in this paper.This corpus comprises different speaking modalities, including scripted text speech, natural conversation, and free speech.The corpus has been compiled in two different languages, viz., French and Spanish.The experimental setups for the recording of the corpus, the acquisition protocols, and the employed equipment are described.Statistics regarding the number and gender of the speakers, number of words, number of sentences, and duration of the recording sessions, are also provided.Preliminary results from the analysis of the correlation among speech, head and hand movements during spontaneous speech are also presented in this paper, showing that acoustic prosodic features are related with head and hand gestures.

Advancing computational thinking in elementary education has been rapidly gaining attention due to the prospective of developing 21st century skills. However, interventions in this domain risk failure if they do not explicitly address the particular socio-cultural traits of the deployment scenario. This is the case in most countries of Latin America, where computing has not reached a sustainable penetration in K-12 education. In order to bridge this gap, we designed a one-week workshop for advancing computational thinking targeted to 10-12 years old Chilean students with no prior experience in programming. This paper describes our intervention and presents the results of a qualitative study analyzing positive and negative aspects of the experience. Although most participants effectively acquired basic programming skills by the end of the intervention, we also identified several inaccurate views on computing and computer scientists. For instance, computing was mostly perceived as a set of informal experiences rather than a way for enabling creation, automation, and work. The word "hacking" appears to be used as a metaphor for more technical terms, such as "programming" or "algorithm". Finally, negative stereotypical views of computer scientists resulting from the intervention were not as frequent as initial perceptions. These results provide fresh evidence on how to design, adapt, and evaluate computational thinking interventions targeted to K-6 students in Latin America.

This work presents and studies the efficiency problem of mapping GPU threads onto simplex domains. A non-linear map$\lambda (\omega)$is formulated based on a block-space enumeration principle that reduces the number of thread-blocks by a factor of approximately$2\times$and$6\times$for 2-simplex and 3-simplex domains, respectively, when compared to the standard approach. Performance results show that$\lambda (\omega)$is competitive and even the fastest map when ran in recent GPU architectures such as the Tesla V100, where it reaches up to$1.5\times$of speedup in 2-simplex tests. In 3-simplex tests, it reaches up to$2.3\times$of speedup for small workloads and up to$1.25\times$for larger ones. The results obtained make$\lambda (\omega)$a useful GPU optimization technique with applications on parallel problems that define all-pairs, all-triplets or nearest neighbors interactions in a 2-simplex or 3-simplex domain.

The study of data-parallel domain re-organization and thread-mapping techniques are relevant topics as they can increase the efficiency of GPU computations on spatial discrete domains with non-box-shaped geometry. In this work we study the potential benefits of applying a succinct data re-organization of a tetrahedral data-parallel domain of size O(n3) combined with an efficient block-space GPU map of the form g (λ) : N → N3. Results from the analysis suggest that in theory the combination of these two optimizations produce significant performance improvement as block-based data reorganization allows a coalesced one-to-one correspondence at local thread-space while g(λ) produces an efficient block-space spatial correspondence between groups of data and groups of threads, reducing the number of unnecessary threads from O(n3) to O(n2ρ3) with ρ ∊ O(1). From the analysis, we obtained that a block based succinct data re-organization can provide up to 2× improved performance over a linear data organization while the map can be up to 6× more efficient than a bounding box approach. The results from this work can serve as a useful guide for a more efficient GPU computation on tetrahedral domains found in spin lattice, finite element and special n-body problems, among others.

In this paper, a system for speech-driven animation of generic 3D head models is presented. The system is based on the inversion of a joint Audio-Visual Hidden Markov Model to estimate the visual information from speech data. Estimated visual speech features are used to animate a simple face model. The animation of a more complex head model is then obtained by automatically mapping the deformation of the simple model to it. The proposed algorithm allows the animation of 3D head models of arbitrary complexity through a simple setup procedure. The resulting animation is evaluated in terms of intelligibility of visual speech through subjective tests, showing a promising performance.

Automated consistency checking of UML models becomes necessary as models grow in size and complexity. Since the UML metamodel does not enforce model consistency, there are no fixed guidelines on how to approach the consistency problem. Current solutions are generally partial. The translation of the metamodel and the user designed model into Description Logics has proved to provide a solution in detecting a large set of inconsistencies. In order to make this solution available to system designers, we have implemented MCC+, a UML model consistency checker, built as a plug-in for Poseidon for UML, and relying on Jena as a reasoning engine. Compared to other approaches, we propose a usable and scalable solution, interoperable with a known modeling tool. We show the application of MCC+ to a real world large example of a meshing tool.

This paper addresses the problem of the allocation of spatial grids for complex nonplanar three-dimensional (3-D) semiconductor device structures. We have characterized the class of meshes suitable for the integration of the device equations with the usual numerical schemes as being a subclass of the class of Delaunay meshes. We propose an algorithm for the efficient generation of such admissible meshes based on the iterative refinement of coarse elements. The generated meshes permit an exact geometrical modeling of rather general domain boundaries of modern silicon devices avoiding the "obtuse angle problem" by construction.>

The authors discuss an automatic mesh generator, Omega , developed as a front-end for the simulation of complex three-dimensional semiconductor devices. Grids generated with Omega exhibit smooth transitions from dense to coarse grid regions and a proper description of irregular geometries such as nonuniform surfaces and interfaces. In addition to the grid itself, Omega provides the dual lattice needed for the integration of the device equations. The underlying algorithm avoids the obtuse angle problem. It is shown how this problem can be formalized in 3-D, how badly shaped elements can be detected, and how they can be avoided by construction. Examples of simulations of complex 3-D devices, with grids with several tens of thousands of mesh points, demonstrate the capabilities of Omega .>