Benchmarking YOLOv8 Variants for Object Detection Efficiency on Jetson Orin NX for Edge Computing Applications
Open accessThe DOI matches its Crossref record; 49 of the 56 references it cites were found in a registry or an index; 2 of the 8 sources its claims cite were checked for support.
Abstract
Edge AI is redefining the deployment of computer vision systems by enabling real-time inference directly on resource-constrained edge devices. This shift offers significant advantages in terms of reduced latency, data privacy, and operational autonomy in bandwidth-limited computing environments. In this paper, we present a systematic performance benchmarking of multiple variants of YOLOv8 on the NVIDIA Jetson Orin NX platform, focusing on object detection tasks. We evaluate inference latency, frame throughput, and computational resource usage across varying input sizes and model complexities. Furthermore, we validate the deployment effectiveness through practical use cases, such as vehicle and package detection. Our findings show that the TensorRT model outperforms PyTorch by 17.7% at a batch size of 2, although PyTorch presents greater stability at larger batch sizes (e.g., 8), where TensorRT encounters resource constraints. In terms of memory usage, it increases linearly with batch size: 69% from batch 1 to 4, with TensorRT requiring 429.20 MB at batch size 2 compared to PyTorch’s 451.24 MB. Furthermore, the processing time per image decreases by 42% when scaling from batch size 1 to 4, highlighting a critical saturation point for edge resources. In summary, the results provide insight into the trade-offs between model size and speed, offering guidance for selecting detection architectures tailored to real-time edge applications.
Overview
CoChat Summary
TL;DR
This paper benchmarks five YOLOv8 variants on an NVIDIA Jetson Orin NX 16 GB for edge object detection, comparing PyTorch and TensorRT across batch sizes, input resolutions, and two use cases. It reports TensorRT speedups, batch-size efficiency gains, memory scaling, and practical stability limits for real-time deployment.
Key findings
- TensorRT outperformed PyTorch by 17.7% at batch size 2, but PyTorch was more stable at larger batch sizes such as 8, where TensorRT encountered resource constraints. For vehicle detection, TensorRT reduced runtime by 52.4–63.3% versus PyTorch; YOLOv8x showed the largest absolute improvement at 36.48 s faster and a 63.3% reduction, while YOLOv8n.engine was fastest at 13.218 s.
- Scaling from batch 1 to 4 in TensorRT-optimized vehicle/package systems improved frame rate by 35.4% to 32.04 FPS and reduced per-frame detection time by 39.9% to 16.27 ms; total execution time fell 26.2% from 12.72 to 9.39 s. GPU memory increased 69% from 25.46 MB to 42.96 MB, with batch 2 showing the lowest memory increase at 429.20 MB versus 453.91 MB at batch 4.
- Vehicle detection showed an accuracy–latency trade-off: YOLOv8x achieved the highest accuracy (0.75) but required 144.636 s, while YOLOv8n was fastest (105.894 s) with the lowest accuracy (0.61). Batch processing improved runtime by approximately 16–21%, with YOLOv8m showing the highest relative improvement (21%).
- For package detection, fine-tuned YOLOv8n achieved 81.5% precision and 72.2% recall across all classes; the package class reached 86.4% precision and 77.3% recall, with mAP@0.5 of 81.7% and mAP@0.5:0.95 of 67.3%.
Methods
- Hardware/software: NVIDIA Jetson Orin NX 16 GB in MAXN (25 W) mode, JetPack 5.1.1 (L4T 35.3.1), CUDA 11.4, TensorRT 8.5; training performed in Google Colab.
- Models/data: five YOLOv8 variants (n/s/m/l/x, 3.0–68.2 M parameters) initialized with COCO pretrained weights, SGD with lr0 = 0.01, 300 epochs, patience 50. Vehicle detection used 2644 frames; package detection used a Roboflow dataset with 12,042 training, 1187 validation, and 16 test images; stratified 80/10/10 split with Mosaic, MixUp, and HSV augmentation.
- Optimization/evaluation: PyTorch weights exported to ONNX (Opset 12), compiled with trtexec to TensorRT engines using FP16; custom Python wrapper with GPU NMS; each run repeated ten times. Metrics included mAP@0.5, end-to-end latency, FPS across batch sizes 1, 2, 4, 8, 16, 32, 64, and 128, and resolutions 640 × 640 and 1200 × 1200, comparing PyTorch versus TensorRT.
Limitations
- The paper states performance metrics came from a limited dataset; real-world variable object densities may yield different latency profiles. The baseline-crossing counting logic assumes generally unidirectional traffic, does not account for vehicles reversing after crossing the line, and may miss counts or switch IDs under significant partial occlusions at the counting line.
- The paper states experiments used a custom active cooling kit; passive cooling enclosures could thermal-throttle earlier and reduce sustained FPS below reported values. Direct comparisons with other edge AI studies are difficult due to hardware configuration variations such as JetPack versions, power modes, and background process overhead.
- Evident from the text: the paper reports conflicting batch-size stability limits—recommending capping TensorRT batch sizes at 4 after OOM failures above batch 8 in one section, while another section recommends capping at 64 after OOM failures above batch 64 for YOLOv8l/x. The abstract’s PyTorch memory figure of 451.24 MB at batch 2 is also not reflected in Table 5.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
1 Valid · 1 Not valid · 0 Not checked · 69 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Specifically, previous works have not systematically identified the batch size regimes at which GPU memory pressure leads to inference engine failures or runtime instability.”
What the sources say
- Not checked — no full text of the sourcecites [3] Alqahtani, D.K., Cheema, M.A., and Toosi, A.N. (2024, January 3–6). Benchmarking deep learning models for object detection on edge computing devices. Proceedings of the 2024 International Conference on Service-Oriented Computing, Tunis, Tunisia. source
- Not checked — no full text of the sourcecites [8] Lazarevich, I., Grimaldi, M., Kumar, R., Mitra, S., Khan, S., and Sah, S. (2023, January 2–3). Yolobench: Benchmarking efficient object detectors on embedded systems. Proceedings of the 2023 IEEE/CVF International Conference on Computer Vision, Paris, France. source
- Not checked — no full text of the sourcecites [26] Feng, H., Mu, G., Zhong, S., Zhang, P., and Yuan, T. (2022). Benchmark analysis of yolo performance on edge intelligence devices. Cryptography, 6. source
- Could not verifycites [51] Rey, L., Bernardos, A.M., Dobrzycki, A.D., Carramiñana, D., Bergesio, L., Besada, J.A., and Casar, J.R. (2025). A performance analysis of you only look once models for deployment on constrained computational edge devices in drone applications. arXiv. sourceExperimental — accuracy not yet measured
Section 3.1.2 mentions a training batch size of 16 limited by GPU hardware, but does not address batch sizes for inference. Sections 2.1 and 4.1 discuss prior edge-device performance studies and a library-related precision limitation, respectively; neither says whether previous works identified batch-size regimes causing memory-related engine failures or runtime instability. Searches for batch size, memory pressure, and inference failures found no passage addressing that claim. (LLM, conf=0.73)
- Not checked — no full text of the sourcecites [52] Mahmud, B.U., Hong, G., Lalwani, V.R., Brown, N., and Asher, Z.D. (2025). Real-time identification of look-alike medical vials using mixed reality-enabled deep learning. Future Internet, 17. source
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The model achieved an overall precision of 81.5% and recall of 72.2% across all classes.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The larger models (YOLOv8l and YOLOv8x) failed to complete processing with 128 batches, indicating memory limits.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Based on our experimental evaluation on the NVIDIA Orin NX 16 GB, we observed consistent Out-of-Memory (OOM) failures when the TensorRT Engine format exceeded batch sizes of 8.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Meanwhile, based on our experimental evaluation on the NVIDIA Orin NX (16 GB), we observed consistent Out-of-Memory (OOM) failures when exceeding batch sizes of 64 for YOLOv8large models (YOLOv8l and YOLOv8x).”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
45 more claims
Create a free account to see all 71 claims and every source verdict.
- It may state two separate claims as one. 18
- Probably not a substantial claim. 5
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 3.2. Data Acquisition and Preprocessing
“We evaluated the models using two distinct use cases, applying a stratified split of 80% for training, 10% for validation, and 10% for testing across our datasets.”
§ 3.2. Data Acquisition and Preprocessing
“This dataset contains 12,042 training images, 1187 validation images, and 16 dedicated test images [53].”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ PHASEI Environment Configuration
“Software Stack: JetPack 5.1.1 (L4T 33.3.1), CUDA 11.4, TensorRT 8.5.”
§ 3.1. Environment Configuration and Hardware Setup
“The software stack was configured using JetPack 5.1.1 (L4T 35.3.1), which provides the necessary drivers and libraries, including CUDA 11.4 and TensorRT 8.5.”
Citations
Cites
You Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803
YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art for Real-Time Object Detectors
Wang, C.-Y., Bochkovskiy, A., & Liao, H.-Y. M. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.00721 · Cited by 13,198
YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications
Li, C., Li, L., Jiang, H., Weng, K., Geng, Y., Li, L., Ke, Z., Li, Q., Cheng, M., Nie, W., et al. (2022) · arXiv · doi:10.48550/arxiv.2209.02976 · Cited by 1,792
A Review on YOLOv8 and Its Advancements
Sohan, M., Sai Ram, T., & Rami Reddy, C. V. (2024) · Algorithms for Intelligent Systems · doi:10.1007/978-981-99-7962-2_39 · Cited by 898
Efficient Acceleration of Deep Learning Inference on Resource-Constrained Edge Devices: A Review
Shuvo, M. M. H., Islam, S. K., Cheng, J., & Morshed, B. I. (2023) · Proceedings of the IEEE · doi:10.1109/jproc.2022.3226481 · Cited by 436
Liang, S., Wu, H., Zhen, L., Hua, Q., Garg, S., Kaddoum, G., Hassan, M. M., & Yu, K. (2022) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2022.3158253 · Cited by 367
Ma, B., Hua, Z., Wen, Y., Deng, H., Zhao, Y., Pu, L., & Song, H. (2024) · Artificial Intelligence in Agriculture · doi:10.1016/j.aiia.2024.02.001 · Cited by 161
TensorRT-Based Framework and Optimization Methodology for Deep Learning Inference on Jetson Boards
Jeong, E., Kim, J., & Ha, S. (2022) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3508391 · Cited by 150
Machine Learning at the Edge: A Data-Driven Architecture With Applications to 5G Cellular Networks
Polese, M., Jana, R., Kounev, V., Zhang, K., Deb, S., & Zorzi, M. (2020) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2020.2999852 · Cited by 131
Benchmark Analysis of YOLO Performance on Edge Intelligence Devices
Feng, H., Mu, G., Zhong, S., Zhang, P., & Yuan, T. (2022) · Cryptography · doi:10.3390/cryptography6020016 · Cited by 100
Cited by
Benchmarking Large Language Model Inference on Limited-Resource Edge Systems
Giedra, H., Matuzevičius, D., Sledević, T., Shubitidze, G., & Serackis, A. (2026) · Electronics · doi:10.3390/electronics15112451 · Cited by 3
Zhang, H.-W., Qin, Y.-M., Wu, A.-Q., Xi, X., Hu, P., & Wang, R.-F. (2026) · Plants · doi:10.3390/plants15081218 · Cited by 13
YOLOv8-EVdet: An emergency vehicle detection method for autonomous vehicles
Loukas, P., Bassir, D., & Amanatiadis, A. A. (2026) · International Journal of Advanced Robotic Systems · doi:10.1177/17298806261486243
Intelligent Surveillance System for Knife Detection Based on YOLOv8 Neural Networks
Peña, L. F. J., & Ibarra, M. J. (2026) · International Multidisciplinary Journal of Emerging Technologies and Applications · doi:10.67294/97s55377
Towards Robust Cross-Dataset Pothole Detection Through Multi-Dataset Pretraining
Widihasaniputri, H. S., Putra, O. V., & Murhadi, M. (2026) · International Journal of Advances in Data and Information Systems · doi:10.59395/ijadis.v7i2.1551
Metadata
- Authors
- Aljami, H. M., Alrowais, N. A., AlAwajy, A. M., Alhrgan, S. O., Aldwaani, R. A., Alsawadi, M. S., Saqib, N. U., Alam, S. S., & Alsubaie, R.
- Affiliations
- King Abdulaziz City for Science and Technology (SA)
- Riyadh Elm University (SA)
- Corresponding author
- Alsawadi, Motasem Samer
- ORCID
- Venue
- Computers
- Publisher
- MDPI AG
- Published
- 2026-02-01
- Volume
- 15
- Issue
- 2
- Pages
- 74
- ISSN
- 2073-431X (ISSN-L)
- OpenAlex
- W7126409579
- Cited by
- 24
- References
- 56
- Citations by year
- 2026: 24
- FWCI
- 68.61
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Computers (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Object Detection Techniques
- Keywords
- YOLOv8, object detection, edge computing, inference latency, memory usage, batch size, TensorRT, PyTorch, vehicle detection
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-09-30, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-09-30 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Benchmarking YOLOv8 Variants for Object Detection Efficiency on Jetson Orin NX for Edge Computing Applications. https://app.cochat.ai/papers/10.3390/computers15020074
- The paper
- Aljami, H. M., Alrowais, N. A., AlAwajy, A. M., Alhrgan, S. O., Aldwaani, R. A., Alsawadi, M. S., Saqib, N. U., Alam, S. S., & Alsubaie, R. (2026). Benchmarking YOLOv8 Variants for Object Detection Efficiency on Jetson Orin NX for Edge Computing Applications. Computers. https://doi.org/10.3390/computers15020074
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.