Dataflow computing has proved to be more efficient for certain high-performance computing algorithms. The prerequisites are that there is enough parallel calculation to cover the overhead of executing instructions on the dataflow hardware until the first result is ready. This is often true with algorithms that work with big data and that can process multiple iterations independently, e.g., while simulating certain phenomena in many elementary volumes. However, dataflow hardware runs typically at an order of magnitude lower frequencies compared to the control-flow processor. From a programmer’s point of view, programming dataflow architectures is considerably harder than programming control-flow architectures. As a result, it is not always obvious whether programming dataflow architectures for certain algorithms is worth the effort needed. Therefore, there is a need for a programmer to be able to predict the outcome of programming for dataflow architectures in terms of accelerating program execution and power savings. This article presents a newly developed tool that a programmer can use for profiling control-flow algorithms and estimating the acceleration possibilities using the dataflow hardware.

Accelerating the solution of linear systems of equations is critical due to their central role in numerous applications, such as numerical simulations, data analytics, and machine learning. This paper presents an analog solver circuit designed to accelerate the solution of symmetric positive definite (SPD) linear systems of equations. The proposed design leverages noninverting operational amplifier configurations to create a negative resistance circuit, effectively modeling any symmetric system. The paper details the principles behind the design, optimizations of the system architecture, and numerical results that demonstrate the robustness of the design. The findings reveal that the proposed system solves symmetric diagonally dominant (SDD) matrices with O(1) complexity, achieving the theoretical maximum speed as the circuit relies solely on resistors. For non-diagonally dominant SPD systems, the solution speed depends on matrix properties, specifically eigenvalues and diagonal dominance deviation, but remains independent of the size of the matrix.

Personality Computing is a field at the intersection of Personality Psychology and Computer Science. Started in 2005, research in the field utilizes computational methods to understand and predict human personality traits. The expansion of the field has been very rapid and, by analyzing digital footprints (text, images, social media, etc.), it helped to develop systems that recognize and even replicate human personality. While offering promising applications in talent recruiting, marketing and healthcare, the ethical implications of Personality Computing are significant. Concerns include data privacy, algorithmic bias, and the potential for manipulation by personality-aware Artificial Intelligence. This paper provides an overview of the field, explores key methodologies, discusses the challenges and threats, and outlines potential future directions for responsible development and deployment of Personality Computing technologies.

This article presents a taxonomy and represents a repository of open problems in computing for numerically and logically intensive problems in a number of disciplines that have to synergize for the best performance of simulation-based feasibility studies on nature-oriented engineering in general and civil engineering in particular. Topics include but are not limited to: Nature-based construction, genomics supporting nature-based construction, earthquake engineering, and other types of geophysical disaster prevention activities, as well as the studies of processes and materials of interest for the above. In all these fields, problems are discussed that generate huge amounts of Big Data and are characterized with mathematically highly complex Iterative Algorithms. In the domain of applications, it has been stressed that problems could be made less computationally demanding if the number of computing iterations is made smaller (with the help of Artificial Intelligence or Conditional Algorithms), or if each computing iteration is made shorter in time (with the help of Data Filtration and Data Quantization). In the domain of computing, it has been stressed that computing could be made more powerful if the implementation technology is changed (Si, GaAs, etc.…), or if the computing paradigm is changed (Control Flow, Data Flow, etc.…).

This article presents and analyses four computing paradigms that are present in today's IT programming world - Control Flow, Data Flow, Diffusion Flow, and Energy Flow. It compares their main properties, points out what purposes each has, and describes what are their advantages and disadvantages. In the third part of this article, the Authors speculate on the possible architecture of a supercomputer on a chip and in the fourth part, they suggest the optimal distribution of resources for a specified set of Civil engineering applications.

This article describes a teaching strategy that synergizes computing and management, aimed at the running of complex projects in industry and academia, in the areas of civil engineering, physics, geosciences, and a number of other related fields. The course derived from this strategy includes four parts: (a) Computing with a selected set of modern paradigms-the stress is on Control Flow and Data Flow computing paradigms, but paradigms conditionally referred to as Energy Flow and Diffusion Flow are also covered; (b) Project management that is holistic-the stress is on the wide plethora of issues spanning from the preparation of project proposals, all the way to incorporation activities to follow after the completion of a successful project; (c) Examples from past research and development experiences-the stress is on experiences of leading experts from academia and industry; (d) Student projects that stimulate creativity-the stress is on methods that educators could use to induce and accelerate the creativity of students in general. Finally, the article ends with selected pearls of wisdom that could be treated as suggestions for further elaboration.

This paper presents a survey on the emerging approaches to the end-to-end congestion control (EECC) in modern Internet. The actual mechanisms are inefficient in the operational contexts of complex and dynamic communication systems. The emerging learning-based approaches employ Machine Learning (ML) to either improve the existing schemes or to completely remodel congestion control from scratch. We classify ML algorithms according to the application area within the context of congestion control, and describe key classes of applications. We also discuss research and engineering issues regarding their implementation and their potential to be included within new operating systems.

This is a short survey of ten algorithms that are often used for military purposes, followed by analysis of their potential suitability for dataflow and GaAs, which are a specific architecture and technology for supercomputers on a chip, respectively. Whenever an algorithm or a device is used in military settings, it is natural to assume strict requirements related to speed, reliability, scale, energy, size, and accuracy. The two aforementioned paradigms seem to be promising in fulfilling most of these requirements.

This article introduces 10 different tensor operations, as well as their generalizations and implementations within the dataflow paradigm. Tensor operations can be utilized for addressing a number of big data problems in machine learning and data mining, such as deep learning, clustering, classification, dimension reduction, anomaly detection, and applications in civil and geo-engineering. With proliferation of data and devices, the main challenge is finding a way to process big quantities of data, especially in environments where resources are limited. This article sheds light on the energy efficient dataflow implementation of tensor operations used in machine learning. The iterative nature of tensor operations and a large amount of data makes them suitable for the dataflow paradigm. All dataflow implementations are analyzed comparatively with the related control-flow implementations, for speedup, complexity, power savings, and meantime between failures. The core contribution of this article is to classify existing implementations of tensor operations on four main architectural approaches in environments with limited resources and to propose energy efficient implementation of 10 tensor operations used in machine learning algorithms. The proposed implementations on the dataflow paradigm are compared against the traditional control-flow paradigm for various data set sizes, and in various conditions of interest.

This article discusses specialized computer cluster architectures for high performance computing that include both control-flow and DataFlow components, as well as their runtime scheduling algorithms. A novel optimal scheduling algorithm for such architectures is proposed. The proposed algorithm is general, but is limited in some cases due to its time complexity. From the base optimal algorithm, two additional heuristic algorithms are derived, and then compared to other schedulers. These heuristic algorithms produce near-optimal schedules for both DataFlow hardware jobs and control-flow jobs at large job counts, with negligible scheduling penalty. Compared to an optimal scheduler, the performance gain decreases slightly as job count increases. This research illustrates that the performance of existing cluster structures can be considerably improved by adding appropriate DataFlow accelerators and a proper scheduling algorithm, while at the same time decreasing the system transistor count and power consumption.

This article describes a course that teaches VLSI design in an unconventional way, to steer creativity and to motivate students to research new computing paradigms, like those presented in this Volume of Advances in Computers.

This paper presents an implementation of Perceptron algorithm using DataFlow approach. DataFlow is a new paradigm that is suitable for solving BigData problems in applications. The iterative nature of the algorithm and a large training set make it suitable for implementing on DataFlow accelerators. Also we discuss differences between ControlFlow paradigm, which is based on the von Neumann architecture, and the DataFlow paradigm, which is based on the Feynman architecture and relies on reconfigurable FPGA chips. We compare these paradigms through aspects such as speed, power dissipation and size. We provide the implementation of Perceptron algorithm using DataFlow hardware accelerator. Results show that DataFlow implementation of Perceptron algorithm is 6 times faster than the implementation on the single core Intel CPU.

This is a short survey of ten algorithms that are often used for military purposes, followed by analysis of their potential suitability for dataflow and GaAs, which are a specific architecture and technology for supercomputers on a chip, respectively. Whenever an algorithm or a device is used in military settings, it is natural to assume strict requirements related to speed, reliability, scale, energy, size, and accuracy. The two aforementioned paradigms seem to be promising in fulfilling most of these requirements.

With the emergence of new technologies, computer networks are becoming more structurally complex, diverse and heterogenous. The increasing discrepancy (among the interconnected networks) in data rates, delays, packet loss, and transmission scenarios, influence significantly the dynamics of congestion control (CC) parametrization. In contrast to the traditional endto-end CC algorithms that rely on strict rules, new approaches aim to involve machine learning in order to continuously adapt the CC to real-time network requirements. However, due to the high computational complexity and memory consumption, the feasibility of these schemes may still be questioned. This paper surveys selected machine-learning based approaches to CC and proposes a roadmap to their implementation in computer systems, by using dataflow computing and Gallium Arsenide (GaAs) chips.

A growing number of research papers shed light on automated machine learning (AutoML) frameworks, which are becoming a promising solution for building complex machine learning models without human expertise and assistance. The key challenge in enabling AutoML frameworks to build an efficient model for anomaly detection tasks is to determine the best underlying model for a given task and optimization metric. The meta-learning approaches based on a set of meta-features that describes data properties can enable efficient model selection in AutoML frameworks. The existing meta-learning approaches based on statistical and information-theoretic meta-features require large amounts of data and computational resources to extract data properties. This paper proposes a novel set of meta-features for model selection in anomaly detection tasks based on domain-specific properties of data which overcomes the shortcomings of existing meta-features by introducing simple but effective meta-features that can be efficiently extracted or estimated by using a low amount of data. Experiments with 63 datasets from different repositories with varying schemas show that the proposed set of meta-features achieves an accuracy of 87% for model selection, while the achieved accuracy for simple meta-features is 74%, for statistical meta-features 68%, for information theory meta-feature 70%, and for a comprehensive set of meta-features by pyMFE 73%. This demonstrates that the proposed set can be adopted by AutoML frameworks across a diverse range of domains.

This paper reviews the most notable works applying machine learning techniques (ML) in the context of geophysics and corresponding subbranches. We showcase both the progress achieved to date as well as the important future directions for further research while providing an adequate background in the fields of weather forecast, wind energy, wave energy, oil and gas exploration. The objective is to reflect on the previous successes and provide a comprehensive review of the synergy between these two fields in order to speed up the novel approaches of machine learning techniques in geophysics. Last but not least, we would like to point out possible improvements, some of which are related to the implementation of ML algorithms using DataFlow paradigm as a means of performance acceleration.

This chapter starts from the assumption that near future 100BTransistor SuperComputers-on-a-Chip will include N big multi-core processors, 1000N small many-core processors, a TPU-like fixed-structure systolic array accelerator for the most frequently used machine learning algorithms needed in bandwidth-bound applications, and a flexible-structure reprogrammable accelerator for less frequently used machine learning algorithms needed in latency-critical applications. The future SuperComputers-on-a-Chip should include effective interfaces to specific external accelerators based on quantum, optical, molecular, and biological paradigms, but these issues are outside the scope of this chapter.

This article starts from the assumption that near future 100BTransistor SuperComputers-on-a-Chip will include N big multi-core processors, 1000N small many-core processors, a TPU-like fixed-structure systolic array accelerator for the most frequently used Machine Learning algorithms needed in bandwidth-bound applications and a flexible-structure reprogrammable accelerator for less frequently used Machine Learning algorithms needed in latency-critical applications.

Contemporary healthcare systems face growing demand for their services, rising costs, and a workforce. Artificial intelligence has the potential to transform how care is delivered and to help meet the challenges. Recent healthcare systems have been focused on using knowledge management and AI. The proposed solution is to reach explainable and causal AI by combining the benefits of the accuracy of deep-learning algorithms with visibility on the factors that are important to the algorithm's conclusion in a way that is accessible and understandable to physicians. Therefore, the authors propose AI approach in which the encoded clinical guidelines and protocols provide a starting point augmented by models that learn from data. The new structure of electronic health records that connects data from wearables and genomics data and innovative extensible big data architecture appropriate for this AI concept is proposed. Consequently, the proposed technology may drastically decrease the need for expensive software and hopefully eliminates the need to do diagnostics in expensive institutions.