Search

The Workflows Community Summit gathered 111 participants from 18 countries to discuss emerging trends and challenges in scientific workflows, focusing on six key areas: time-sensitive workflows, AI-HPC convergence, multi-facility workflows, heterogeneous HPC environments, user experience, and FAIR computational workflows. The integration of AI and exascale computing has revolutionized scientific workflows, enabling higher-fidelity models and complex, time-sensitive processes, while introducing challenges in managing heterogeneous environments and multi-facility data dependencies. The rise of large language models is driving computational demands to zettaflop scales, necessitating modular, adaptable systems and cloud-service models to optimize resource utilization and ensure reproducibility. Multi-facility workflows present challenges in data movement, curation, and overcoming institutional silos, while diverse hardware architectures require integrating workflow considerations into early system design and developing standardized resource management tools. The summit emphasized improving user experience in workflow systems and ensuring FAIR workflows to enhance collaboration and accelerate scientific discovery. Key recommendations include developing standardized metrics for time-sensitive workflows, creating frameworks for cloud-HPC integration, implementing distributed-by-design workflow modeling, establishing multi-facility authentication protocols, and accelerating AI integration in HPC workflow management. The summit also called for comprehensive workflow benchmarks, workflow-specific UX principles, and a FAIR workflow maturity model, highlighting the need for continued collaboration in addressing the complex challenges posed by the convergence of AI, HPC, and multi-facility research environments.

Recording the provenance of scientific computation results is key to the support of traceability, reproducibility and quality assessment of data products. Several data models have been explored to address this need, providing representations of workflow plans and their executions as well as means of packaging the resulting information for archiving and sharing. However, existing approaches tend to lack interoperable adoption across workflow management systems. In this work we present Workflow Run RO-Crate, an extension of RO-Crate (Research Object Crate) and Schema.org to capture the provenance of the execution of computational workflows at different levels of granularity and bundle together all their associated objects (inputs, outputs, code, etc.). The model is supported by a diverse, open community that runs regular meetings, discussing development, maintenance and adoption aspects. Workflow Run RO-Crate is already implemented by several workflow management systems, allowing interoperable comparisons between workflow runs from heterogeneous systems. We describe the model, its alignment to standards such as W3C PROV, and its implementation in six workflow systems. Finally, we illustrate the application of Workflow Run RO-Crate in two use cases of machine learning in the digital image analysis domain. A corresponding RO-Crate for this article is at https://w3id.org/ro/doi/10.5281/zenodo.10368989, Comment: 38 pages, 5 figures, 3 tables. Resubmitted to PLOS ONE following peer review

In the present article we study the periodic structure of some well-known classes of $C^1$ self-maps on the product of spheres of different dimensions: transversal maps, Morse-Smale diffeomorphisms and maps with all its periodic points hyperbolic. Our approach is via the Lefschetz fixed point theory. We give a complete characterization of the minimal set of Lefschetz periods for Morse-Smale diffeomorphisms on these spaces. We also consider $C^1$ maps with all its periodic points hyperbolic and we give conditions for these maps to have infinitely many periodic points. We describe the period set of the transversal maps on these spaces. Finally we applied these results to describe the periodic structure of similar classes of maps on compact, connected, simple connected and simple Lie groups.

We consider spaces of multivariate splines defined on a particular type of simplicial partitions that we call (generalized) oranges. Such partitions are composed of a finite number of maximal faces with exactly one shared medial face. We reduce the problem of finding the dimension of splines on oranges to computing dimensions of splines on simpler, lower-dimensional partitions that we call projected oranges. We use both algebraic and Bernstein-B\'ezier tools., Comment: 18 pages, 3 figures

In this work, we study the thermalization between two bodies separated by a vacuum gap by coupling the non-Fourier behavior of the materials with the radiative heat transfer in the near-field. Unlike the diffusion-type temperature profile, in non-Fourier materials, the temperature behaves as a wave, changing the thermalization process. Due to the temperature profile induced by the coupling with conduction, we show that the radiative heat flux exchanged between the two bodies differs from the Fourier case, and exhibits transient temperature effects at the onset of the thermalization process. These results have important implications in nanoscale thermal management, near-field solid-state cooling, and nanoscale energy conversion., Comment: 5 figures

One of the signatures of superconductivity is the formation of the Abrikosov vortex lattice in type-II superconductors in the presence of an external magnetic field. Here, we study the near-field radiative heat transfer between a spherical nanoparticle and a nearby planar substrate, both made of optimally-doped YBa$_2$Cu$_3$O$_{7-\delta}$. We show that the heat flux displays a periodic spatial pattern congruent with the material optical response modulated by the Abrikosov lattice. Our results enrich the toolbox of methods available to study non-conventional superconductivity., Comment: 3 figures, Supplementary material available

Scientific workflows have become integral tools in broad scientific computing use cases. Science discovery is increasingly dependent on workflows to orchestrate large and complex scientific experiments that range from execution of a cloud-based data preprocessing pipeline to multi-facility instrument-to-edge-to-HPC computational workflows. Given the changing landscape of scientific computing and the evolving needs of emerging scientific applications, it is paramount that the development of novel scientific workflows and system functionalities seek to increase the efficiency, resilience, and pervasiveness of existing systems and applications. Specifically, the proliferation of machine learning/artificial intelligence (ML/AI) workflows, need for processing large scale datasets produced by instruments at the edge, intensification of near real-time data processing, support for long-term experiment campaigns, and emergence of quantum computing as an adjunct to HPC, have significantly changed the functional and operational requirements of workflow systems. Workflow systems now need to, for example, support data streams from the edge-to-cloud-to-HPC enable the management of many small-sized files, allow data reduction while ensuring high accuracy, orchestrate distributed services (workflows, instruments, data movement, provenance, publication, etc.) across computing and user facilities, among others. Further, to accelerate science, it is also necessary that these systems implement specifications/standards and APIs for seamless (horizontal and vertical) integration between systems and applications, as well as enabling the publication of workflows and their associated products according to the FAIR principles. This document reports on discussions and findings from the 2022 international edition of the Workflows Community Summit that took place on November 29 and 30, 2022.

We numerically study the tuning of the radiative heat transfer between a spherical InSb nanoparticle in the vicinity of a flat SiC surface assisted by a static magnetic field. By changing the value of the applied magnetic field, the dielectric function of the nanosphere becomes anisotropic due to the excitation of magneto-plasmons. In the dipolar approximation, the plasmon resonance of the particle splits into two additional satellite resonances that shift to higher and lower frequencies as the field increases. When one of the particle resonances overlaps with the phonon-polariton frequency of the SiC surface, an enhancement of the heat transfer of two orders of magnitude is obtained. To understand the tuning of the radiative heat transfer, we present a detailed analysis of the nature of the modes that can be excited (surface, bulk, and hyperbolic).

In this paper we solve the Cattaneo-Vernotte Equation for a periodic heterostructure made of alternate layers of different materials. The solutions describe thermal waves traveling in a periodic system, and it allows us to introduce the concept of thermal crystals. We show that the dispersion relation shows the characteristics of a band-structure, however the corresponding Bloch wave vector is always complex corresponding to pseudo-bands, unlike what happens in photonic or acoustic crystals. In this context, we also discuss the use of the Floquet-Bloch theorem for thermal waves. The case of finite layered structures is also analyzed showing the possibility of changing the temperature and heat flux by introducing defects opening the possibility of thermal management through the pseudo-band structure., Comment: 5 figures

We present a theoretical calculation of the changes in the Hamaker constant due to the anomalous reduction of the static dielectric function of water. Under confinement, the dielectric function of water decreases from a bulk value of 79 down to 2. If the confining walls are made of a dielectric material, the Hamaker constant reduces almost by 90\%. However, if the confinement is realized with metallic plates, there is little change in the Hamaker constant. Additionally, we show that confinement can be used to decreases the Debye screening length without changing the salt concentration. This, in turn, is used to change the Hamaker constant in the presence of electrolytes., Comment: 6 figures

The simulation of the behavior of the human brain is one of the most ambitious challenges today with a non-end of important applications. We can find many different initiatives in the USA, Europe and Japan which attempt to achieve such a challenging target. In this work, we focus on the most important European initiative (the Human Brain Project) and on one of the models developed in this project. This tool simulates the spikes triggered in a neural network by computing the voltage capacitance on the neurons' morphology, being one of the most precise simulators today. In the present work, we have evaluated the use of MPI+OpenMP tasking on top of this framework. We prove that this approach is able to achieve a good scaling even when computing a relatively low workload (number of neurons) per node. One of our targets consists of achieving not only a highly scalable implementation, but also to develop a tool with a high degree of abstraction without losing control and performance by using \emph{MPI+OpenMP} tasking. The main motivation of this work is the evaluation of this cutting-edge simulation on multi-morphology neural networks. The simulation of a high number of neurons, which are completely different among them, is an important challenge. In fact, in the multi-morphology simulations, we find an important unbalancing between the nodes, mainly due to the differences in the neurons, which causes an important under-utilization of the available resources. In this work, the authors present and evaluate mechanisms to deal with this and reduce the time of this kind of simulations considerably.

Data Envelopment Analysis (DEA) is widely used as a benchmarking tool for improving performance of organizations. For that purpose, DEA analyses provide information on both target setting and peer identification. However, the identification of peers is actually a by-product of DEA. DEA models seek a projection point of the unit under evaluation on the efficient frontier of the production possibility set (PPS), which is used to set targets, while peers are identified simply as the members of the so-called reference sets, which consist of the efficient units that determine the projection point as a combination of them. In practice, the selection of peers is crucial for the benchmarking, because organizations need to identify a peer group in their sector or industry that represent actual performances from which to learn. In this paper, we argue that DEA benchmarking models should incorporate into their objectives criteria for the selection of suitable benchmarks among peers, in addition to considering the setting of appropriate targets (as usual). Specifically, we develop models having two objectives: setting the closest targets and selecting the most similar reference sets. Thus, we seek to establish targets that require the less effort from organizations for their achievement in addition to identifying peer groups with the most similar performances, which are potential benchmarks to emulate and improve.

In benchmarking, organizations look outward to examine others' performance in their industry or sector. Often, they can learn from the best practices of some of them and improve. In order to develop this idea within the framework of Data Envelopment Analysis (DEA), this paper extends the common benchmarking framework proposed in Ruiz and Sirvent (2016) to an approach based on the benchmarking of decision making units (DMUs) against several reference sets. We refer to this approach as cross-benchmarking. First, we design a procedure aimed at making a selection of reference sets (as defined in DEA), which establish the common framework for the benchmarking. Next, benchmarking models are formulated which allow us to set the closest targets relative to the reference sets selected. The availability of a wider spectrum of targets may offer managers the possibility of choosing among alternative ways for improvements, taking into account what can be learned from the best practices of different peer groups. Thus, cross-benchmarking is a flexible tool that can support a process of future planning while considering different managerial implications.

Data Envelopment Analysis (DEA) is extended to the evaluation of performance of organizations within the framework of the implementation of plans for improvements that set management goals. Managers usually set goals without having any evidence that they will be achievable at the moment of conducting performance evaluation or, on the contrary, they may set little too unambitious goals. Using DEA for the benchmarking ensures an evaluation in terms of targets that both are attainable and represent best practices. In addition, the approach we propose adjusts the DEA benchmarking to the goals in order to consider the policy of improvements that was pursued with the setting of such goals. From the methodological point of view, the models that minimize the distance to the DEA strong efficient frontier are extended to incorporate goal information. Specifically, the models developed seek DEA targets that are as close as possible to both actual performances and management goals. To illustrate, we examine an example that is concerned with the evaluation of performance of public Spanish universities.

We present a theoretical study of the modification of the near-field radiative heat transfer due to phonon-plasmon coupling in bilayer systems made of a doped semiconductor and a polar dielectric. By tuning the surface-plasmon mode of the former material in resonance with the phonon modes of the latter one, the near-field spectral radiative heat transfer can be suppressed in a frequency band gap. We distinguish between the interlayer mode coupling within each bilayer and the intragap one between the two bilayers. We elucidate the role of each of them in the formation of the band gap and we determine the distance range at which one mechanism dominates over the other. Furthermore, we show that the surface plasmon polariton of the top-most layer allows optimizing the total heat transfer as a function of the separation of the two bilayers., Comment: 5 figures

Incentive plans involve payments for performance relative to some set of goals. In this paper, we extend Data Envelopment Analysis (DEA) to the evaluation of performance in the specific context of pay-for-performance incentive plans. The approach proposed ensures that the evaluation of performance of decision making units (DMUs) that follow the implementation of incentive plans, is made in terms of targets that are attainable, as well as representing best practices. A model is developed that adjusts the benchmarking to the goals through the corresponding payment of incentives, thus DEA targets are established taking into consideration the improvement strategies that were set out in the incentive plans. To illustrate, we examine an application concerned with the financing of public Spanish universities.

Sellami and Sirvent conjectured that the balanced pair algorithm fails for the following pair of Pisot substitutions: \[ \varphi_0: \begin{array}{l} a \mapsto abc b \mapsto a c \mapsto ac \end{array} \quad \text{ and } \quad \varphi_1: \begin{array}{l} a \mapsto cba b \mapsto a c \mapsto ca \end{array}. \] The conjecture stated the balanced pair algorithm fails because there is no initial balanced pair. In the present note we prove this conjecture using a method based on simultaneous coding of the pair of the fixed points of the morphisms., Comment: 11 pages