The advent of Deep Learning (DL) has radically transformed the computing industry across the entire spectrum from algorithms to circuits. As myriad application domains embrace DL, it has become synonymous with a genre of workloads across vision, speech, language, recommendations, robotics, and games. The key compute kernel within most DL workloads is general matrix-matrix multiplications (GEMMs), which appears frequently during both the forward pass (inference and training) and backward pass (training). GEMMs are a natural choice for hardware acceleration to speed up training, and have led to 2D systolic architectures like NVIDIA tensor cores and Google Tensor Processing Unit (TPU). Unfortunately, emerging GEMMs in DL are highly irregular and sparse, which lead to poor data mappings on systolic architectures. This paper proposes SIGMA, a flexible and scalable architecture that offers high utilization of all its processing elements (PEs) regardless of kernel shape and sparsity. Within SIGMA includes a novel reduction tree microarchitecture named Forwarding Adder Network (FAN). SIGMA performs 5.7x better than systolic array architectures for irregular sparse matrices, and roughly 3x better than state-of-the-art sparse accelerators. We demonstrate an instance of SIGMA operating at 10.8 TFLOPS efficiency across arbitrary levels of sparsity, with a 65.10 mm^2 and 22.33 W footprint on a 28 nm process.
Mujtaba Ali, Muhammad Yaqoob Javed, Aamer Bilal Asghar
et al.
ABSTRACT Increasing global energy demand and renewable energy expansion have heightened the importance of accurate solar irradiance forecasting for effective grid management and capacity planning. Atmospheric pollution significantly affects solar irradiance measurements, requiring air quality integration for precise forecasting in polluted urban environments. This study develops a comprehensive multi‐city data set spanning eight geographically diverse locations with systematically categorized pollution levels, from pristine environments (Copenhagen, Sydney) to heavily polluted urban centers (Beijing, New Delhi, Lahore). A pollution‐aware neural network training methodology is introduced, representing the first systematic investigation of ensemble model performance across explicitly categorized atmospheric quality levels. The study presents a novel ensemble architecture integrating multi‐layer perceptrons, recurrent neural networks, and nonlinear autoregressive with exogenous inputs, specifically designed for forecasting under varying atmospheric pollution conditions. The proposed ensemble model achieves superior performance with R² of 0.8702, RMSE of 1.0809, and MAE of 0.8137, consistently outperforming individual models across all pollution categories and geographical locations. Validation using the HI‐SEAS data set confirms superiority over three contemporary state‐of‐the‐art methodologies. The framework incorporates SHapley Additive exPlanations (SHAP) analysis for model interpretability and comprehensive cross‐validation procedures. This study establishes a foundational framework for pollution‐aware solar forecasting, addressing critical gaps regarding atmospheric variability's impact on prediction accuracy.
Aiming at the problems of intensified air port competition conflicts and uncontrollable transmission delay caused by the dense deployment of multiple access points in the fiber-to-the-room (FTTR) C-WAN2.0 architecture, a collaboration-enhanced distributed channel access (Co-EDCA) mechanism with multi-AP was proposed. The interference domain isolation was achieved through conflict graph modeling and time slot allocation, and the competition window was dynamically adjusted using reinforcement learning to optimize the competitive behavior of multiple SFUs. Simulation results show that the proposed mechanism can reduce the conflict probability by 23%, improve the system throughput by 98%, and reduce the transmission delay by 57% in SFU-dense scenarios, effectively solving the resource competition and quality of service guarantee problems under the C-WAN2.0 architecture.
In this study, a performance comparison experiment with a vortex exhaust installed at the end of a ventilation device to enhance the effect induced by reducing indoor pollutants was conducted. The experiment was carried out by constructing a mock-up room with a limited indoor environment, and performances were compared based on the following two tests. First, to confirm the effect of pollutant reduction, the wind speed was measured based on the distance from each exhaust system to verify the depth and speed at which wind can flow. Pollutants were induced to the vortex exhaust, general exhaust gasses were generated, and their performances were compared. Second, Arizona dust was used to confirm the performance with regard to the removal of pollutants which existed in particulate form (PM 10), and for CO<sub>2</sub> gas, a representative gaseous pollutant was used as a reference. Based on the results, it was confirmed that installing a vortex exhaust system can allow for the generation of wind speeds that allow propagation at greater depths (>110 mm) compared to cases in which general exhaust is used; accordingly, exhaust performance can be achieved at increased depths. In addition, the experiment confirmed that vortex exhaust can improve the efficiency of simultaneous removal of PM 10 and CO<sub>2</sub> compared with general exhaust. Further, it was shown that installing a vortex exhaust system can remove PM 10 and CO<sub>2</sub> farther from the exhaust port in a shorter period than a general exhaust port. In addition, it was inferred that vortex exhaust can be utilized to prevent indoor pollutants and diseases in combination with the latest technology.
How can AI help us connect to the past in terms of conservation? How can 17-year-old photos be helpful in renewed preservation efforts? This research aims to use AI to connect both in a seamless 3D reconstruction of heritage from images taken of Gongfan Palace, Yunlin, Taiwan. AI-assisted 3D modeling was used to reconstruct the details of these images across different 3D platforms of the 3DGS or NeRF models generated by Postshot<sup>®</sup>, RODIN<sup>®</sup>, and KIRI Engine<sup>®</sup>. Mesh and point models created using Zephyr<sup>®</sup> were referred to and assessed in three sets. The consistent and inconsistent reconstructed results also included AI-assisted modeling outcomes in Stable Diffusion<sup>®</sup>- and Postshot<sup>®</sup>-based animations, followed by a 3D assessment and section-based composition analysis. The AI-assisted environment concluded with a recursive reconstruction involving 3D models and 2D images. AI assisted the 3D modeling process in an alternative approach, producing extraordinary structural and visual details. AI-trained models can be assessed and their use extended to composition analysis by section. Evolved documentation and interpretation using AI enables new structures and the management of resources, formats, and interfaces as part of continuous preservation efforts.
The purpose of this study is to explore the architecture and functioning of hybrid solar desalination systems and investigate their potential as a sustainable solution for water purification. The study reveals that solar-powered desalination systems offer a remarkable alternative to traditional methods, as they rely on clean solar energy and produce no noise or sound pollution. In addition, they have demonstrated cost-effectiveness in generating drinking water, especially in desert regions and inaccessible areas. Furthermore, the research highlights the significance of incorporating waste heat energy into the desalination process. Also shows that utilizing waste heat energy can significantly reduce expenses and enhance the overall effectiveness of water desalination. Through an in-depth analysis of the fundamental principles and real-world applications, this study underscores the importance and rationale for implementing hybrid solar desalination systems. By effectively utilizing solar energy, these systems provide a sustainable approach to address water scarcity and ensure the efficient management of water and energy resources. This study emphasizes the fundamental importance of the structure of hybrid solar desalination systems fueled by solar energy in the efficient management of water resources. By combining technological innovations with renewable energy sources, these systems pave the way for a sustainable future.
HIGHLIGHTS
Hybrid solar desalination systems, which rely on solar energy as their major power source for purifying water.;
This review paper explores the architecture and functioning of hybrid solar desalination systems.;
This review paper emphasizes the significance and rationale for utilizing hybrid solar desalination systems that rely on solar energy to efficiently handle water and energy resources.;
Tanımı ve sınırları olan bir yerleşke, üniversite kampüsü olarak yüksek öğretim ve araştırma için yaygın kullanılan bir yerleşim tipolojisidir. Bulunduğu bölgeden sınırları ile ayrılan bu yerleşkeler içlerindeki yapı ve açık alanları ile eğitim-öğretim vermektedir. Şehir veya kırsal alanlar ile ulaşım ilişkilerini ve görsel temaslarını farklı ölçeklerdeki giriş kapıları ile sağlarlar. Her kapı eşik ve sınır kavramını farklı yorumlar. Türkiye’de, özellikle 2000 yılı sonrasında çok sayıda üniversite kurulmuş, yerleşke planlaması ile birlikte dışarıyla ilişki kuracak kapılarının tasarımı ve inşası yapılmıştır. Kapı tasarımlarında; güvenlik, ulaşım düzenleme ve eşik tanımlama gibi fonksiyonları ile birlikte üniversitenin dışarıya bakan ilk yapısı olması nedeniyle anlamsal içeriği de etkili olmuştur. Bu kapıları derleyen ve üzerinden değerlendirme yapan az sayıda çalışma vardır. Bunlardan bazıları tasarım yaklaşımlarını sınıflandırmaya çalışmıştır. Bu çalışmada 2010 yılında kurulan Erzurum Teknik Üniversitesi’nin Giriş Kapısı üzerinden form ve malzeme arayışı anlatılmış, tasarım yaklaşımları ile birlikte tartışılmıştır. Formun ön planda olduğu ve bir referansa dayanan bu kapıda, işlevlerin ve geometri tasarımının referans biçim ile ilişki kurma yöntemleri incelenmiştir. Çalışmada sunulan yaklaşımlar ve örnek üzerinden değerlendirilen uygulama, form ve malzeme seçimleri, giriş kapıları tasarım yaklaşımları alanına katkı sunmaktadır.
This work proposes a smart crop growth monitoring system that contains an adaptive cryptography engine to ensure the security of sensor data and an edge artificial intelligence (AI) based estimator to classify the pest and disease severity (PDS) of target crops. Based on the smart system management mechanism, cryptographic functions can be adapted to varying and real-time requirements, while the actuators can be controlled to interact with the physical world to ensure the healthy growth of crops. Experiments show when all the four cryptographic hardware modules, including RTEA32, RTEA64, XTEA32 and XTEA64, are supported, using the adaptive cryptography engine, 72.4% of slice LUTs and 68.4% of slice registers in terms of the Xilinx Zynq-7000 XC7Z020 chip can be saved. Through the smart system management mechanism, a power consumption of 0.009 watts can be reduced. Furthermore, using the binarized neural network (BNN) hardware module of the PDS estimator, the recognition accuracy of target crops i.e. dragon fruits can achieve 76.57%. Compared to the microprocessor-based design and the GPU accelerated one, the same BNN architecture on the FPGA can accelerate the frames per second by a factor of 4,919.29 and a factor of 1.08, respectively.
Performance evaluation is crucial for environmental design and sustainable development, especially so for architecture and landscape architecture. However, such performance evaluations remain rare in practice. It is argued that the concerns over potential negative evaluations and a lack of funding are the two main barriers preventing the undertaking of performance evaluations. This research investigated how these two barriers were overcome in practice by studying 41 evaluation cases in the New Zealand landscape architecture field, as well as several international and architectural case studies for comparison. A range of enablers for performance evaluation practices were identified by this research, including funding sources and models that were not documented by existing literature, as well as two strategies for handling the risks of negative evaluation. All of the identified enablers share the same underlying logic—the benefits and costs of an evaluation should be well-regulated by certain mechanisms to keep the benefits of an evaluation greater than, or at least balanced with, the costs, for all the parties involved in the evaluation.
Viktor J Radermacher, Vincent Fernandez, Emma R Schachner
et al.
Ornithischian dinosaurs were ecologically prominent herbivores of the Mesozoic Era that achieved a global distribution by the onset of the Cretaceous. The ornithischian body plan is aberrant relative to other ornithodiran clades, and crucial details of their early evolution remain obscure. We present a new, fully articulated skeleton of the early branching ornithischian Heterodontosaurus tucki. Phase-contrast enhanced synchrotron data of this new specimen reveal a suite of novel postcranial features unknown in any other ornithischian, with implications for the early evolution of the group. These features include a large, anteriorly projecting sternum; bizarre, paddle-shaped sternal ribs; and a full gastral basket – the first recovered in Ornithischia. These unusual anatomical traits provide key information on the evolution of the ornithischian body plan and suggest functional shifts in the ventilatory apparatus occurred close to the base of the clade. We complement these anatomical data with a quantitative analysis of ornithischian pelvic architecture, which allows us to make a specific, stepwise hypothesis for their ventilatory evolution.
To fundamentally solve the bottleneck of Von Neumann’s computing architecture, a neuromorphic thin-film transistor (NTFT) employing Pb(Zr, Ti)O<sub>3</sub> (PZT) was investigated. The indium gallium zinc oxide (IGZO) channel back gate TFT structure was chosen to solve the diffusion of atoms that form a channel layer during the annealing process for crystallization of PZT. A post-deposition process with IGZO after annealing PZT and using an oxide-based material as a channel structure can minimize the diffusion phenomenon of junction materials and oxygen together, which leads to a high and reliable performance of the NTFT. The basic operations of synapses short-term memory (STM) and long-term memory (LTM) were also analyzed to confirm the application of a neuromorphic device. The high dielectric constant and polarization properties of Pb(Zr, Ti)O<sub>3</sub> (PZT) allow the power consumption of spike signals used in spike dependent plasticity change to be reduced to 10 pJ. Moreover, a wide dynamic range of <inline-formula> <tex-math notation="LaTeX">$\text{G}_{\mathrm {max}}/\text{G}_{\mathrm {min}} \cong ~1000$ </tex-math></inline-formula> was obtained, and the channel conductance was maintained over 40000 seconds. The optimized pulse achieved multi-level states (>32), which made the learning process efficient. This study verified that the PZT-TFT structure has a high potential and merits for neuromorphic devices.
In this paper, we demonstrate that there are vortex beam solutions for the photon in the rotating medium. By constructing the photon wave function with Riemann–Silberstein vector, we derive the dynamic equation of the photon in moving medium from the Maxwell equations and the non-relativistic Minkowski relations. In case of the stationary state, the dynamic equation of the photon can be written as a Dirac-like equation, where the velocity of the medium plays the role of a vector potential. By giving the medium different forms of rotating velocity fields, we obtain different vortex beam solutions of the photon, such as the diffracting and non-diffracting Laguerre–Gaussian (LG) beam solutions via proper approximations. For the diffracting LG beam solution, we acquire a new term arising from the medium rotation that can change the Gouy phase, and then accordingly predict the rotation behavior of the photon interference pattern. In addition, the rotation of the medium can lead to the change of the relative intensity distribution of the interference pattern. Furthermore, our theory predicts the existence of the Landau levels of transverse photon energy in the nondiffracting LG beam solution.