Implementation of Thermal Event Image Processing Algorithms on NVIDIA Tegra Jetson TX2 Embedded System-on-a-Chip

Jabłoński, Bartłomiej; Makowski, Dariusz; Perek, Piotr

doi:10.3390/en14154416

Cited by 6 publications

(12 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These two edge computers are currently being used in several fields such as manufacturing, agriculture, retail, life sciences, etc. For instance, an image processing algorithm for thermal events has been recently proposed for the Jetson TX2 [14]. The configurations of both devices are shown in Table I To assess the energy profile of these edge computers, we measure the timing and energy figures of neural network architectures, focusing on one single layer of the whole NN architecture.…”

Section: Methodsmentioning

confidence: 99%

Energy Consumption of Neural Networks on NVIDIA Edge Boards: an Empirical Model

Lahmer,

Khoshsirat,

Rossi

et al. 2022

Preprint

View full text Add to dashboard Cite

Recently, there has been a trend of shifting the execution of deep learning inference tasks toward the edge of the network, closer to the user, to reduce latency and preserve data privacy. At the same time, growing interest is being devoted to the energetic sustainability of machine learning. At the intersection of these trends, in this paper we focus on the energetic characterization of machine learning at the edge, which is attracting increasing attention. Unfortunately, calculating the energy consumption of a given neural network during inference is complicated by the heterogeneity of the possible underlying hardware implementation. In this work, we aim at profiling the energetic consumption of inference tasks for some modern edge nodes by deriving simple but accurate models. To this end, we performed a large number of experiments to collect the energy consumption of fully connected and convolutional layers on two well-known edge boards by NVIDIA, namely, Jetson TX2 and Xavier. From these experimental measurements, we have then distilled a simple and practical model that can provide an estimate of the energy consumption of a certain inference task on these edge computers. We believe that this model can prove useful in many contexts as, for instance, to guide the search for efficient neural network architectures, as a heuristic in neural network pruning, to find energy-efficient offloading strategies in a split computing context, or to evaluate and compare the energy performance of deep neural network architectures.

show abstract

Section: Methodsmentioning

confidence: 99%

Energy Consumption of Neural Networks on NVIDIA Edge Boards: an Empirical Model

Lahmer,

Khoshsirat,

Rossi

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The paper focuses on the evaluation of performance and some accuracy aspects of the developed image processing system for plasma diagnostics presented in Figure 2. Overload hotspot and surface layer detection algorithms were described and benchmarked in [16]. Correction and calibration algorithms, i.e., Non-Uniformity Correction (NUC), Bad Pixel Correction (BPC) and thermal calibration, are not examined in the paper since they shall be executed on a Field-Programmable Gate Array (FPGA) for the highest performance due to their simplicity.…”

Section: Research Objectivementioning

confidence: 99%

“…The details regarding the selection of C++/CUDA algorithms, implementation and optimisation are described in the following paragraphs. The further minor characteristics of datasets, scene models and software dependencies are described in [16].…”

Section: Infrared Image Processingmentioning

confidence: 99%

“…Blob analysis is performed in parallel for source and destination PFCs. Blob tracking is performed with the correspondence criteria that match blobs between consecutive frames by evaluating the relative change in overlap and area [16,36]. The GPU Block-based Union Find (BUF) algorithm [37] is used for Connected Component Labelling (CCL).…”

Section: Reflection Detectionmentioning

confidence: 99%

See 1 more Smart Citation

Evaluation of NVIDIA Xavier NX Platform for Real-Time Image Processing for Plasma Diagnostics

et al. 2022

Self Cite

View full text Add to dashboard Cite

Machine protection is a core task of real-time image diagnostics aiming for steady-state operation in nuclear fusion devices. The paper evaluates the applicability of the newest low-power NVIDIA Jetson Xavier NX platform for image plasma diagnostics. This embedded NVIDIA Tegra System-on-a-Chip (SoC) integrates a Graphics Processing Unit (GPU) and Central Processing Unit (CPU) on a single chip. The hardware differences and features compared to the previous NVIDIA Jetson TX2 are signified. Implemented algorithms detect thermal events in real-time, utilising the high parallelism provided by the embedded General-Purpose computing on Graphics Processing Units (GPGPU). The performance and accuracy are evaluated on the experimental data from the Wendelstein 7-X (W7-X) stellarator. Strike-line and reflection events are primarily investigated, yet benchmarks for overload hotspots, surface layers and visualisation algorithms are also included. Their detection might allow for automating real-time risk evaluation incorporated in the divertor protection system in W7-X. For the first time, the paper demonstrates the feasibility of complex real-time image processing in nuclear fusion applications on low-power embedded devices. Moreover, GPU-accelerated reference processing pipelines yielding higher accuracy compared to the literature results are proposed, and remarkable performance improvement resulting from the upgrade to the Xavier NX platform is attained.

show abstract

“…Since the publication of article [16], the world has made great progress in hightemperature superconductor technology (necessary for plasma stabilization) and in the work on controlled thermonuclear fusion. Some of these works have been described in publications [20][21][22][23][24][25][26][27][28][29][30][31]. In [22], the use of electric energy from fusion in transport in Europe is modeled.…”

Section: Introductionmentioning

confidence: 99%

Feasibility Study of Low Mass and Low Energy Consumption Drilling Devices for Future Space (Mining Surveying) Missions

et al. 2021

View full text Add to dashboard Cite

The global climate crisis forces the search for new ecological sources of energy and mining methods. Space mining can solve those problems, but, first, wide geological surveying space missions using drilling methods are necessary. Additionally, drilling methods will be important in geological, life searching, geoengineering, and many other studies of extraterrestrial objects. Space is becoming a new area of possible drilling applications. Designing future space drilling missions requires adapting drilling technologies, not only to the conditions of the space environment, but also to the economic and technological realities of the space industry. The possibility of constructing low mass coring devices with energy consumption below 100 W was investigated in this paper. Minimization of energy consumption and mass of a coring is essential for the device to be used in space missions, when lander instruments supplied by low power electric battery are expected to work reliably and the launch cost (depending of mass) at an economically acceptable level. Some similar devices investigated for the future space missions are known from papers listed in the references. To answer whether or not it is possible to build such devices, the authors performed initial drillability tests. The obtained results are presented in this paper.

show abstract

Implementation of Thermal Event Image Processing Algorithms on NVIDIA Tegra Jetson TX2 Embedded System-on-a-Chip

Cited by 6 publications

References 24 publications

Energy Consumption of Neural Networks on NVIDIA Edge Boards: an Empirical Model

Energy Consumption of Neural Networks on NVIDIA Edge Boards: an Empirical Model

Evaluation of NVIDIA Xavier NX Platform for Real-Time Image Processing for Plasma Diagnostics

Feasibility Study of Low Mass and Low Energy Consumption Drilling Devices for Future Space (Mining Surveying) Missions

Contact Info

Product

Resources

About