ABSTRACT As an essential part of traditional Chinese handicrafts, Qinghai embroidery embodies rich cultural connotations and unique artistic value. However, with the development of modern society, traditional handicrafts face severe challenges in inheritance and protection. To effectively address these challenges and promote Qinghai embroidery's digital safety and inheritance, this study realizes the automatic classification and identification of Qinghai embroidery images based on the SE‐ResNet152V2 model. First, we constructed an image dataset containing five kinds of Qinghai embroidery patterns, including Tu nationality Pan embroidery, Huangzhong Dui embroidery, Hehuang embroidery, Mongolian embroidery, and Tibetan Guinan embroidery. The regions that contribute the most when the model judges the image categories are revealed by the GRAD‐CAM technique, and the data are preprocessed by the image enhancement technique to enhance the data diversity and improve the model's generalization ability. For the complexity and detailed features of Qinghai embroidery patterns, this paper introduces the squeeze‐and‐excitation (SE) attention module to enhance the model's ability to capture key features. By systematically comparing the effects of multiple optimizers and attention mechanisms combination models, the optimal combination of the Nadam optimizer and SE attention mechanism is finally selected. The experimental results show that the accuracy of the optimized SE‐ResNet152V2 model on the self‐built Qinghai embroidery image dataset is 91.73%, which is 11.43% higher than that of the original ResNet152V2 model. Further experiments show that the SE‐ResNet152V2 model is better than other popular neural network models such as MobileNetV1, EfficientNetB2, vision transformer (VIT), and swin transformer, regarding classification accuracy. It has proven its effectiveness and superiority in processing Qinghai embroidery pattern recognition tasks and provided strong technical support for digital protection and inheritance of traditional crafts.

For a complex trait, heritability ([Formula: see text]) gives the genetic determination of its variation. Given the emergence of biobank-scale data, a more powerful method is needed to estimate [Formula: see text]. Based on the framework of Haseman-Elston regression (RHE-reg), we integrate a fast randomization algorithm to estimate [Formula: see text], and RHE-reg can tackle biobank-scale data, such as UK Biobank (UKB), very efficiently. Furthermore, we present an analytical solution that balances computational cost and precision of the estimation, a property that is important in dealing with biobank-scale data. We investigated the performance of the RHE-reg in simulated data and also applied it for 81 UKB quantitative traits; as tested in UKB data of nearly 300,000 unrelated individuals, it took on average about 4.5 hours to complete an estimation when used 10 CPUs. We extended the application of RHE-reg into distributed datasets when privacy is not compromised. As shown in UKB and simulated data the performance of RHE-reg was accurate in estimating [Formula: see text]. The software for estimating SNP-heritability for biobank-scale data is released.

Due to the end of Moore’s Law and Dennard Scaling, Domain-Specific Accelerators (DSAs) have come to a Cambrian explosion. Especially when advancing into the intelligent era, more and more DSAs are integrated into System-on-Chips (SoCs) as intellectual property (IP) blocks to provide high performance and efficiency. Currently, IPs usually expose IP-dependent hardware interfaces, requiring SoCs to manage them as isolated devices with software running on the host CPU. However, such software-managed heterogeneity in CPU-centric SoCs leads to low IP utilization. This inefficiency arises from the dependence on software optimization, coupled with the control and data exchange overheads. To improve IP utilization of heterogeneous SoCs, in this article, we rearchitect the SoC as a processor (i.e., SaaP) to orchestrate hardware heterogeneity. SaaP features an orchestration pipeline where DSAs are integrated as execution units and managed directly by the hardware pipeline to conceal the hardware heterogeneity from software. Moreover, SaaP redesigns the register file and data paths to implement an IP-level data-forwarding mechanism, avoiding the costly control and data exchange in the CPU-centric execution model. Block data dependence among different DSAs is carefully resolved to exploit mixed-level parallelism and inter-IP data exchange. SaaP abstracts tasks as mixed-scale instructions, where each instruction can be mapped to different IPs. Experimental results show that compared against Xavier on six fully software-optimized benchmarks from different domains, SaaP-rearchitected Xavier achieves a$2.08{\times }$speedup, with an 8.21% area reduction and only 2.98% increase in power consumption.

Relightable 4D avatar reconstruction which enables high fidelity and real-time rendering continues to be a crucial but challenging problem, especially from monocular videos. Previous NeRF-based 4D avatars enable photo-realistic relighting but are too slow for rendering, while point-based or mesh-based 4D avatars are efficient but have limited rendering quality. The recent success of 3D Gaussian Splatting, i.e., 3DGS, has inspired a series of impressive 4D Gaussian avatars, however, most of which only focus on faithful appearance reconstruction but are not relightable. To address such issues, this article proposes a new Relightable 4D Gaussian Avatar, i.e., RGAvatar, tailored for high fidelity relightable rendering from monocular videos. Our key idea is to introduce a new relightable 4D Gaussian representation, based on which we can directly perform high fidelity Physically Based Rendering, and an effective joint learning mechanism for compact 4D Gaussian reconstruction with SDF regulation and accurate materials and lighting decomposition. By comparing with previous state-of-the-art approaches, RGAvatar can significantly outperform previous approaches in relightable rendering quality and speed. To our best knowledge, RGAvatar contributes a new state-of-the-art 4D Gaussian avatar from monocular videos, which enables high fidelity relightable rendering in a quite efficient manner.

With the development of artificial intelligence and the improvement of hardware computing power, deep learning models have become widely used in the Internet of Things (IoT) field, especially for analyzing spatiotemporal data collected by wireless sensors. Recurrent neural networks (RNNs) such as long short-term memory (LSTM) network are generally used for these time-series data. Hyperparameter settings of model training are regarded as essential factors for the performance of deep learning models. Manually optimizing hyperparameters not only cost more resources but also be more likely to set hyperparameters that follow stereotypes, resulting in unreasonable hyperparameter settings and poor model performance. As one of the most important fields in automated machine learning research, automated hyperparameter optimization (HPO) mainly includes grid search, hyperparameter search based on genetic algorithm, etc. Whereas these methods have their own drawbacks. In this article, an automated HPO method based on parallel genetic algorithm (PGA) is proposed. According to the process of PGA, this article divided HPO into several stages, including population initialization, fitness function, tournament selection, crossover operators, mutation operators, subgroup exchange, and end of evolution. Then, the proposed HPO method is implemented in LSTM models and tested on two different time-series datasets collected by real-world IoT sensors. By comparing our proposed method with other mainstream HPO methods in different datasets, it is proved that our HPO method based on PGA shows a better performance on both time costs and prediction results.

Graph neural networks (GNNs), which extend traditional neural networks for processing graph-structured data, have been widely used in many fields. The GNN computation mainly consists of theedge processingto generate messages by combining the edge/vertex features and thevertex processingto update the vertex features with aggregated messages. In addition to nontrivial vector operations in the edge processing, huge random accesses and neural network operations in the vertex processing, the graph topology of GNNs may also vary during the computation (i.e., dynamic GNNs). The above characteristics pose significant challenges on existing architectures. In this article, we propose a novel accelerator named CAMBRICON-G for efficient processing of both dynamic and static GNNs. The key of CAMBRICON-G is to abstract the irregular computation of a broad range of GNN variants to the process of regularly tiledadjacent cuboid(which extends the traditional adjacent matrix of graph by adding the dimension of vertex features). The intuition is that the adjacent cuboid facilitates exploitation of both data locality and parallelism by offeringmultidimensional multilevel tiling(including spatial and temporal tiling) opportunities. To perform themultidimensional spatial tiling, the CAMBRICON-G architecture mainly consists of the cuboid engine (CE) and hybrid on-chip memory. The CE has multiple vertex processing units (VPUs) working in a coordinated manner to efficiently process the sparse data and dynamically update the graph topology with dedicated instructions. The hybrid on-chip memory contains the topology-aware cache and multiple scratchpad memory to reduce off-chip memory access. To perform themultidimensional temporal tiling, an easy-to-use programming model is provided to flexibly explore different tiling options for large graphs. Experimental results show that compared against Nvidia P100 GPU, the performance and energy efficiency can be improved by$7.14\times $and$20.18\times $, respectively, on various GNNs, which validates both the versatility and energy efficiency of CAMBRICON-G.

Machine learning techniques are pervasive tools for emerging commercial applications and many dedicated machine learning computers on different scales have been deployed in embedded devices, servers, and data centers. Currently, most machine learning computer architectures still focus on optimizing performance and energy efficiency instead of programming productivity. However, with the fast development in silicon technology, programming productivity, including programming itself and software stack development, becomes the vital reason instead of performance and power efficiency that hinders the application of machine learning computers. In this article, we propose Cambricon-F, which is a series of homogeneous, sequential, multi-layer, layer-similar, and machine learning computers with same ISA. A Cambricon-F machine has a fractal von Neumann architecture to iteratively manage its components: it is with von Neumann architecture and its processing components (sub-nodes) are still Cambricon-F machines with von Neumann architecture and the same ISA. Since different Cambricon-F instances with different scales can share the same software stack on their common ISA, Cambricon-Fs can significantly improve the programming productivity. Moreover, we address four major challenges in Cambricon-F architecture design, which allow Cambricon-F to achieve a high efficiency. We implement two Cambricon-F instances at different scales, i.e., Cambricon-F100 and Cambricon-F1. Compared to GPU based machines (DGX-1 and 1080Ti), Cambricon-F instances achieve 2.82x, 5.14x better performance, 8.37x, 11.39x better efficiency on average, with 74.5, 93.8 percent smaller area costs, respectively. We further propose Cambricon-FR, which enhances the Cambricon-F machine learning computers to flexibly and efficiently support all the fractal operations with a reconfigurable fractal instruction set architecture. Compared to the Cambricon-F instances, Cambricon-FR machines achieve 1.96x, 2.49x better performance on average. Most importantly, Cambricon-FR computers are able to save the code length with a factor of 5.83, thus significantly improving the programming productivity.

In this paper, we propose a novel pairwise crowd-sourcing model to reduce the uncertainty of top-k ranking using a crowd of domain experts. Given a crowdsourcing task of limited budget, we propose efficient algorithms to select the best object pairs for crowdsourcing that will bring in the highest quality improvement. Extensive experiments show that our proposed solutions outperform a random selection method by up to 30 times in terms of quality improvement of probabilistic top-kranking queries. In terms of efficiency, our proposed solutions can reduce the elapsed time of a brute-force algorithm from several days to one minute.

Probabilistic top-k ranking is an important and well-studied query operator in uncertain databases. However, the quality of top-k results might be heavily affected by the ambiguity and uncertainty of the underlying data. Uncertainty reduction techniques have been proposed to improve the quality of top-k results by cleaning the original data. Unfortunately, most data cleaning models aim to probe the exact values of the objects individually and therefore do not work well for subjective data types, such as user ratings, which are inherently probabilistic. In this paper, we propose a novel pairwise crowdsourcing model to reduce the uncertainty of top-k ranking using a crowd of domain experts. Given a crowdsourcing task of limited budget, we propose efficient algorithms to select the best object pairs for crowdsourcing that will bring in the highest quality improvement. Extensive experiments show that our proposed solutions outperform a random selection method by up to 30 times in terms of quality improvement of probabilistic top-k ranking queries. In terms of efficiency, our proposed solutions can reduce the elapsed time of a brute-force algorithm from several days to one minute.

Subgraph query (via subgraph isomorphism) is a fundamental and powerful query in various real graph applications. It has actively been investigated for performance enhancements recently. However, due to the high complexity of subgraph query, hosting efficient subgraph query services has been a technically challenging task, because the owners of graph data may not always possess the IT expertise to offer such services and hence may outsource to query service providers (SP). SPs are often equipped with high performance computing utilities (e.g., a cloud) that offer better scalability, elasticity and IT management. Unfortunately, as SPs may not always be trusted, security (such as the confidentiality of messages exchanged) has been recognized as one of the critical attributes of Quality of Services (QoS) [4]. This influences the willingness of both data owners and query clients to use SP's services. Recently, there is a bloom on the research on query processing with privacy preservation1, e.g., in the context of relational databases, spatial databases and graph databases. However, up to date, private subgraph query has not yet been studied.

Online learning is used in a wide range of real applications, e.g., predicting ad click-through rates (CTR) and personalized recommendations. Based on the analysis of users' behaviors in Video-On-Demand (VoD) recommender systems,we discover that the most recent users' actions can better reflect users' current intentions and preferences. Under this observation, we thereby propose a novel time-decaying online learning algorithm derived from the state-of-the-art FTRL-proximal algorithm, called Time-Decaying Adaptive Prediction (TDAP) algorithm.

One fundamental type of query for graph databases is subgraph isomorphism queries (a.k.a subgraph queries). Due to the computational hardness of subgraph queries coupled with the cost of managing massive graph data, outsourcing the query computation to a third-party service provider has been an economical and scalable approach. However, confidentiality is known to be an important attribute of Quality of Service (QoS) in Query as a Service (QaaS). In this paper, we propose the first practical private approach for subgraph query services, asymmetric structure-preserving subgraph query processing, where the data graph is publicly known and the query structure/topology is kept secret. Unlike other previous methods for subgraph queries, this paper proposes a series of novel optimizations that only exploit graph structures, not the queries. Further, we propose a robust query encoding and adopt the novel cyclic group based encryption so that query processing is transformed into a series of private matrix operations. Our experiments confirm that our techniques are efficient and the optimizations are effective.

Subgraph queries have been a fundamental query for retrieving patterns from graph data. Due to the well known NP hardness of subgraph queries, those queries may sometimes take a long time to complete. Our recent investigation on real- world datasets revealed that the performance of queries on graphs generally varies greatly. In other words, query clients may occasionally encounter “unexpectedly” long execution from a subgraph query processor. This paper aims to demonstrate a tool that alleviates the problem by monitoring subgraph query progress. Specifically, we present a novel subgraph query progress indicator called PIGEON that exploits query-time information to report to users accurate estimated query progress. In the demonstration, users may interact with PIGEON to gain insights on the query evaluation, which include the following: Users are enabled to (i) monitor query progress; (ii) analyze the causes of long query times; and (iii) abort queries that run abnormally long, which may sometimes contain human errors.

Keys for graphs aim to uniquely identify entities represented by vertices in a graph. We propose a class of keys that are recursively defined in terms of graph patterns, and are interpreted with subgraph isomorphism. Extending conventional keys for relations and XML, these keys find applications in object identification, knowledge fusion and social network reconciliation. As an application, we study the entity matching problem that, given a graph G and a set Σ of keys, is to find all pairs of entities (vertices) in G that are identified by keys in Σ. We show that the problem is intractable, and cannot be parallelized in logarithmic rounds. Nonetheless, we provide two parallel scalable algorithms for entity matching, in MapReduce and a vertex-centric asynchronous model. Using real-life and synthetic data, we experimentally verify the effectiveness and scalability of the algorithms.

A fundamental problem of graph databases is subgraph isomorphism query (a.k.a subgraph query): given a query graph Q and a graph database, it retrieves the graphs Gs from the database that contain Q. Due to the cost of managing massive data coupled with the computational hardness of subgraph isomorphism testing, outsourcing the computations to a third-party provider is an appealing alternative. However, confidentiality has been a critical attribute of quality of service (QoS) in query services. To the best of our knowledge, subgraph query services with tunable preservation of privacy of structural information have never been addressed. In this paper, we present the first work on structure-preserving subIso (SPsubIso). A crucial step of our work is to transform subIso-the seminal subgraph isomorphism algorithm (the Ullmann's algorithm)-into a series of matrix operations. We propose a novel cyclic group based encryption (CGBE) method for private matrix operations. We propose a protocol that involves the query client and static indexes to optimize SPsubIso. We prove that the structural information of both Q and G are preserved under CGBE and analyze the privacy preservation in the presence of the optimizations. Our extensive experiments on both real and synthetic datasets verify that SPsubIso is efficient and the optimizations are effective.

Subgraph similarity search is used in graph databases to retrieve graphs whose subgraphs are similar to a given query graph. It has been proven successful in a wide range of applications including bioinformatics and chem-informatics, etc. Due to the cost of providing efficient similarity search services on ever-increasing graph data, database outsourcing is apparently an appealing solution to database owners. Unfortunately, query service providers may be untrusted or compromised by attacks. To our knowledge, no studies have been carried out on the authentication of the search. In this paper, we propose authentication techniques that follow the popular filtering-and-verification framework. We propose an authentication-friendly metric index called GMTree. Specifically, we transform the similarity search into a search in a graph metric space and derive small verification objects (VOs) to-be-transmitted to query clients. To further optimize GMTree, we propose a sampling-based pivot selection method and an authenticated version of MCS computation. Our comprehensive experiments verified the effectiveness and efficiency of our proposed techniques.

Graphs are powerful tools suitable for a large variety of applications including chemical databases and the Semantic Web, among others. A fundamental query of graph databases is the subgraph query: given a query graph$q$, it retrieves the data graphs from a database that contain$q$. Due to the cost of managing massive data coupled with the computational hardness of subgraph query processing, outsourcing the processing to a third-party service provider is an appealing alternative. However, security properties such as data integrity and the response time are critical Quality of Service (QoS) issues in query services. Unfortunately, to our knowledge, authenticated subgraph query services have not been addressed before. To support the service, we propose Merkle${\tt IFTree}$$({\tt MIFTree})$where Merkle hash trees are applied into our Intersection-aware Feature-subgraph Tree$({\tt IFTree})$.${\tt IFTree}$aims to minimize I/O in a well-received subgraph query paradigm called the filtering-and-verification framework. The structures that need to be introduced to verification objects (${\cal VO}$s) and the authentication time are both minimized. Subsequently, the overall response time is minimized. For optimizations, we propose an enhanced authentication method on${\tt MIFTree}$. Our detailed experiments on both real and synthetic datasets demonstrate that${\tt MIFTree}$is clearly more efficient than a baseline method.

The Lattice Boltzmann method (LBM) for visual simulation of fluid flow generally employs cubic Cartesian (CC) lattices such as the D3Q13 and D3Q19 lattices for the particle transport. However, the CC lattices lead to suboptimal representation of the simulation space. We introduce the face-centered cubic (FCC) lattice, fD3Q13, for LBM simulations. Compared to the CC lattices, the fD3Q13 lattice creates a more isotropic sampling of the simulation domain and its single lattice speed (i.e., link length) simplifies the computations and data storage. Furthermore, the fD3Q13 lattice can be decomposed into two independent interleaved lattices, one of which can be discarded, which doubles the simulation speed. The resulting LBM simulation can be efficiently mapped to the GPU, further increasing the computational performance. We show the numerical advantages of the FCC lattice on channeled flow in 2D and the flow-past-a-sphere benchmark in 3D. In both cases, the comparison is against the corresponding CC lattices using the analytical solutions for the systems as well as velocity field visualizations. We also demonstrate the performance advantages of the fD3Q13 lattice for interactive simulation and rendering of hot smoke in an urban environment using thermal LBM.

Abstract Due to its high performance/cost ratio, a GPU cluster is an attractive platform for large scale general‐purpose computation and visualization applications. However, the programming model for high performance general‐purpose computation on GPU clusters remains a complex problem. In this paper, we introduce the Zippy frame‐work, a general and scalable solution to this problem. It abstracts the GPU cluster programming with a two‐level parallelism hierarchy and a non‐uniform memory access (NUMA) model. Zippy preserves the advantages of both message passing and shared‐memory models. It employs global arrays (GA) to simplify the communication, synchronization, and collaboration among multiple GPUs. Moreover, it exposes data locality to the programmer for optimal performance and scalability. We present three example applications developed with Zippy: sort‐last volume rendering, Marching Cubes isosurface extraction and rendering, and lattice Boltzmann flow simulation with online visualization. They demonstrate that Zippy can ease the development and integration of parallel visualization, graphics, and computation modules on a GPU cluster.

We simulate 3D fluid flow by a locally-refined lattice Boltzmann method (LBM) on graphics hardware. A low resolution LBM simulation running on a coarse grid models global flow behavior of the entire domain with low consumption of computational resources. For regions of interest where small visual details are desired, LBM simulations are performed on fine grids, which are separate grids superposed on the coarse one. The flow properties on boundaries of the fine grids are determined by the global simulation on the coarse grid. Thus, the locally refined fine-grid simulations follow the global fluid behavior, and model the desired small-scale and turbulent flow motion with their denser numerical discretization. A fine grid can be initiated and terminated at any time while the global simulation is running. It can also move inside the domain with a moving object to capture small-scale vortices caused by the object. Besides the performance improvement due to the adaptive simulation, the locally-refined LBM is suitable for acceleration on contemporary graphics hardware (GPU), since it involves only local and linear computations. Therefore, our approach achieves fast and adaptive 3D flow simulation for computer games and other interactive applications.

We describe a novel volumetric global illumination framework based on the Face-Centered Cubic (FCC) lattice. An FCC lattice has important advantages over a Cartesian lattice. It has higher packing density in the frequency domain, which translates to better sampling efficiency. Furthermore, it has the maximal possible kissing number (equivalent to the number of nearest neighbors of each site), which provides optimal 3D angular discretization among all lattices. We employ a new two-pass (illumination and rendering) global illumination scheme on an FCC lattice. This scheme exploits the angular discretization to greatly simplify the computation in multiple scattering and to minimize illumination information storage. The GPU has been utilized to further accelerate the rendering stage. We demonstrate our new framework with participating media and volume rendering with multiple scattering, where both are significantly faster than traditional techniques with comparable quality.

We provide a physically-based framework for simulating the natural phenomena related to heat interaction between objects and the surrounding air. We introduce a heat transfer model between the heat source objects and the ambient flow environment, which includes conduction, convection, and radiation. The heat distribution of the objects is represented by a novel temperature texture. We simulate the thermal flow dynamics that models the air flow interacting with the heat by a hybrid thermal lattice Boltzmann model (HTLBM). The computational approach couples a multiple-relaxation-time LBM (MRTLBM) with a finite difference discretization of a standard advection-diffusion equation for temperature. In heat shimmering and mirage, the changes in the index of refraction of the surrounding air are attributed to temperature variation. A nonlinear ray tracing method is used for rendering. Interactive performance is achieved by accelerating the computation of both the MRTLBM and the heat transfer, as well as the rendering on contemporary graphics hardware (GPU).

Inspired by the attractive Flops/dollar ratio and the incredible growth in the speed of modern graphics processing units (GPUs), we propose to use a cluster of GPUs for high performance scientific computing. As an example application, we have developed a parallel flow simulation using the lattice Boltzmann model (LBM) on a GPU cluster and have simulated the dispersion of airborne contaminants in the Times Square area of New York City. Using 30 GPU nodes, our simulation can compute a 480x400x80 LBM in 0.31 second/step, a speed which is 4.6 times faster than that of our CPU cluster implementation. Besides the LBM, we also discuss other potential applications of the GPU cluster, such as cellular automata, PDE solvers, and FEM.

We present a system for simulating and visualizing the propagation of dispersive contaminants with an application to urban security. In particular, we simulate airborne contaminant propagation in open environments characterised by sky-scrapers and deep urban canyons. Our approach is based on the multiple relaxation time lattice Boltzmann model (MRTLBM), which can efficiently handle complex boundary conditions such as buildings. In addition, we model thermal effects on the flow field using the hybrid thermal MRTLBM. Our approach can also accommodate readings from various sensors distributed in the environment and adapt the simulation accordingly. We accelerate the computation and efficiently render many buildings with small textures on the GPU. We render streamlines and the contaminant smoke with self-shadowing composited with the textured buildings.

We present an approach for simulating the natural dynamics that emerge from the interaction between a flow field and immersed objects. We model the flow field using the Lattice Boltzmann Model (LBM) with boundary conditions appropriate for moving objects and accelerate the computation on commodity graphics hardware (GPU) to achieve real-time performance. The boundary conditions mediate the exchange of momentum between the flow field and the moving objects resulting in forces exerted by the flow on the objects as well as the back-coupling on the flow. We demonstrate our approach using soap bubbles and a feather. The soap bubbles illustrate Fresnel reflection, reveal the dynamics of the unseen flow field in which they travel, and display spherical harmonics in their undulations. Our simulation allows the user to directly interact with the flow field to influence the dynamics in real time. The free feather flutters and gyrates in response to lift and drag forces created by its motion relative to the flow. Vortices are created as the free feather falls in an otherwise quiescent flow.

We introduce a method for the animation of fire propagation and the burning consumption of objects represented as volumetric data sets. Our method uses a volumetric fire propagation model based on an enhanced distance field. It can simulate the spreading of multiple fire fronts over a specified isosurface without actually having to create that isosurface. The distance field is generated from a specific shell volume that rapidly creates narrow spatial bands around the virtual surface of any given isovalue. The complete distance field is then obtained by propagation from the initial bands. At each step multiple fire fronts can evolve simultaneously on the volumetric object. The flames of the fire are constructed from streams of particles whose movement is regulated by a velocity field generated with the hardware-accelerated Lattice Boltzmann Model (LBM). The LBM provides a physically-based simulation of the air flow around the burning object. The object voxels and the splats associated with the flame particles are rendered in the same pipeline so that the volume data with its external and internal structures can be displayed along with the fire.