{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,7]],"date-time":"2026-02-07T11:01:55Z","timestamp":1770462115652,"version":"3.49.0"},"reference-count":32,"publisher":"MDPI AG","issue":"2","license":[{"start":{"date-parts":[[2026,2,6]],"date-time":"2026-02-06T00:00:00Z","timestamp":1770336000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Future Internet"],"abstract":"<jats:p>AI-driven network security relies increasingly on Large Language Models (LLMs) to detect sophisticated threats; however, their deployment on resource-constrained edge devices is severely hindered by immense parameter scales. While unstructured pruning offers a theoretical reduction in model size, commodity Graphics Processing Unit (GPU) architectures fail to efficiently leverage element-wise sparsity due to the mismatch between fine-grained pruning patterns and the coarse-grained parallelism of Tensor Cores, leading to latency bottlenecks that compromise real-time analysis of high-volume security telemetry. To bridge this gap, we propose SPARTA (Sparse Parallel Architecture for Real-Time Threat Analysis), an algorithm\u2013architecture co-design framework. Specifically, we integrate a hardware-based address remapping interface to enable flexible row-offset access. This mechanism facilitates a novel graph-based column vector merging strategy that aligns sparse data with Tensor Core parallelism, complemented by a pipelined execution scheme to mask decoding latencies. Evaluations on Llama2-7B and Llama2-13B benchmarks demonstrate that SPARTA achieves an average speedup of 2.35\u00d7 compared to Flash-LLM, with peak speedups reaching 5.05\u00d7. These findings indicate that hardware-aware microarchitectural adaptations can effectively mitigate the penalties of unstructured sparsity, providing a viable pathway for efficient deployment in resource-constrained edge security.<\/jats:p>","DOI":"10.3390\/fi18020088","type":"journal-article","created":{"date-parts":[[2026,2,6]],"date-time":"2026-02-06T16:32:59Z","timestamp":1770395579000},"page":"88","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["SPARTA: Sparse Parallel Architecture for Real-Time Threat Analysis for Lightweight Edge Network Defense"],"prefix":"10.3390","volume":"18","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-7454-3320","authenticated-orcid":false,"given":"Shi","family":"Li","sequence":"first","affiliation":[{"name":"College of Cryptology and Cyber Science, Nankai University, Tianjin 300350, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiyun","family":"Mi","sequence":"additional","affiliation":[{"name":"College of Cryptology and Cyber Science, Nankai University, Tianjin 300350, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-3642-5862","authenticated-orcid":false,"given":"Lin","family":"Zhang","sequence":"additional","affiliation":[{"name":"Chinese Institute of New Generation Artificial Intelligence Development Strategies, Nankai University, Tianjin 300350, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ye","family":"Lu","sequence":"additional","affiliation":[{"name":"College of Cryptology and Cyber Science, Nankai University, Tianjin 300350, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,2,6]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"4059","DOI":"10.1109\/TII.2021.3088938","article-title":"FLEAM: A Federated Learning Empowered Architecture to Mitigate DDoS in Industrial IoT","volume":"18","author":"Li","year":"2021","journal-title":"IEEE Trans. Ind. Inform."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"860","DOI":"10.1109\/TII.2020.2974520","article-title":"Deep Learning-based DDoS-attack Detection for Cyber\u2013Physical System over 5G Network","volume":"17","author":"Hussain","year":"2020","journal-title":"IEEE Trans. Ind. Inform."},{"key":"ref_3","first-page":"1877","article-title":"Language Models are Few-Shot Learners","volume":"33","author":"Brown","year":"2020","journal-title":"Adv. Neural Inf. Process. Syst. (NeurIPS)"},{"key":"ref_4","unstructured":"Yang, A., Xiao, B., Wang, B., Zhang, B., Bian, C., Yin, C., Lv, C., Pan, D., Wang, D., and Yan, D. (2023). Baichuan 2: Open Large-scale Language Models. arXiv."},{"key":"ref_5","unstructured":"Chiang, W.-L., Li, Z., Lin, Z., Sheng, Y., Wu, Z., Zhang, H., Zheng, L., Zhuang, S., Zhuang, Y., and Gonzalez, J.E. (2026, January 27). Vicuna: An Open-Source Chatbot Impressing GPT-4 with 90%* ChatGPT Quality, LMSYS Blog, Available online: https:\/\/lmsys.org\/blog\/2023-03-30-vicuna\/."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Lin, X., Xiong, G., Gou, G., Li, Z., Shi, J., and Yu, J. (2022, January 25\u201329). ET-BERT: A contextualized datagram representation with pre-training transformers for encrypted traffic classification. Proceedings of the ACM Web Conference 2022, Virtual.","DOI":"10.1145\/3485447.3512217"},{"key":"ref_7","unstructured":"NVIDIA (2023). NVIDIA Ada GPU Architecture Whitepaper, NVIDIA Corporation. Technical Report."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Song, Y., Mi, Z., Xie, H., and Chen, H. (2024, January 4\u20136). PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU. Proceedings of the 30th Symposium on Operating Systems Principles (SOSP), Austin, TX, USA.","DOI":"10.1145\/3694715.3695964"},{"key":"ref_9","unstructured":"Sheng, Y., Zheng, L., Yuan, B., Li, Z., Ryabinin, M., Chen, B., Liang, P., R\u00e9, C., Stoica, I., and Zhang, C. (2023, January 23\u201329). FlexGen: High-throughput Generative Inference of Large Language Models with a Single GPU. Proceedings of the International Conference on Machine Learning (ICML), Honolulu, HI, USA."},{"key":"ref_10","unstructured":"Touvron, H., Lavril, T., Izacard, G., Martinet, X., Lachaux, M.A., Lacroix, T., Rozi\u00e8re, B., Goyal, N., Hambro, E., and Azhar, F. (2023). Llama: Open and Efficient Foundation Language Models. arXiv."},{"key":"ref_11","unstructured":"Frantar, E., and Alistarh, D. (2023, January 23\u201329). SparseGPT: Massive Language Models Can be Accurately Pruned in One-Shot. Proceedings of the International Conference on Machine Learning (ICML). PMLR, Honolulu, HI, USA."},{"key":"ref_12","unstructured":"Sun, M., Liu, Z., Bair, A., and Kolter, J.Z. (2024, January 7\u201311). A Simple and Effective Pruning Approach for Large Language Models. Proceedings of the Twelfth International Conference on Learning Representations (ICLR), Vienna, Austria."},{"key":"ref_13","unstructured":"Xu, P., Shao, W., Chen, M., Tang, S., Zhang, K., Gao, P., An, F., Qiao, Y., and Luo, P. (2024, January 7\u201311). BESA: Pruning Large Language Models with Blockwise Parameter-Efficient Sparsity Allocation. Proceedings of the Twelfth International Conference on Learning Representations (ICLR), Vienna, Austria."},{"key":"ref_14","unstructured":"Ma, X., Fang, G., and Wang, X. (2023, January 10\u201316). LLM-Pruner: On the Structural Pruning of Large Language Models. Proceedings of the Advances in Neural Information Processing Systems (NeurIPS), New Orleans, LA, USA."},{"key":"ref_15","unstructured":"Ashkboos, S., Croci, M.L., do Nascimento, M.G., Hoefler, T., and Hensman, J. (2024, January 7\u201311). SliceGPT: Compress Large Language Models by Deleting Rows and Columns. Proceedings of the Twelfth International Conference on Learning Representations (ICLR), Vienna, Austria."},{"key":"ref_16","unstructured":"Fan, R., Wang, W., and Chu, X. (May, January 27). DTC-SpMM: Bridging the Gap in Accelerating General Sparse Matrix Multiplication with Tensor Cores. Proceedings of the 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), La Jolla, CA, USA."},{"key":"ref_17","unstructured":"Xia, H., Zheng, Z., Li, Y., Zhuang, D., Zhou, Z., Qiu, X., Li, Y., Lin, W., and Song, S.L. (September, January 28). Flash-LLM: Enabling Cost-Effective and Highly-Efficient Large Generative Model Inference with Unstructured Sparsity. Proceedings of the VLDB Endowment, Vancouver, BC, Canada."},{"key":"ref_18","unstructured":"Hong, K., Dai, G., Xu, J., Mao, Q., Li, X., Liu, J., Chen, K., Dong, Y., and Wang, Y. (2024, January 13\u201316). FlashDecoding++: Faster Large Language Model Inference on GPUs. Proceedings of the Machine Learning and Systems (MLSys), Santa Clara, CA, USA."},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Wang, Y., Zhang, C., Xie, Z., Guo, C., Liu, Y., and Leng, J. (2021, January 14\u201318). Dual-Side Sparse Tensor Core. Proceedings of the 2021 ACM\/IEEE 48th Annual International Symposium on Computer Architecture (ISCA), Virtual.","DOI":"10.1109\/ISCA52012.2021.00088"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Xia, M., Zhong, Z., and Chen, D. (2022, January 22\u201327). Structured Pruning Learns Compact and Accurate Models. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (ACL), Dublin, Ireland.","DOI":"10.18653\/v1\/2022.acl-long.107"},{"key":"ref_21","unstructured":"An, Y., Zhao, X., Yu, T., Tang, M., and Wang, J. (2024, January 20\u201327). Fluctuation-Based Adaptive Structured Pruning for Large Language Models. Proceedings of the Thirty-Eighth AAAI Conference on Artificial Intelligence, Vancouver, BC, Canada."},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Markidis, S., Der Chien, S.W., Laure, E., Peng, I.B., and Vetter, J.S. (2018, January 21\u201325). Nvidia tensor core programmability, performance & precision. Proceedings of the 2018 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), Vancouver, BC, Canada.","DOI":"10.1109\/IPDPSW.2018.00091"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"29","DOI":"10.1109\/MM.2021.3061394","article-title":"NVIDIA A100 Tensor Core GPU: Performance and Innovation","volume":"41","author":"Choquette","year":"2021","journal-title":"IEEE Micro"},{"key":"ref_24","unstructured":"Liu, S., Lindholm, J.E., Siu, M.Y., Coon, B.W., and Oberman, S.F. (2010). Operand Collector Architecture. (No. 7,834,881), U.S. Patent."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Khairy, M., Shen, Z., Aamodt, T.M., and Rogers, T.G. (June, January 30). Accel-Sim: An Extensible Simulation Framework for Validated GPU Modeling. Proceedings of the 2020 ACM\/IEEE 47th Annual International Symposium on Computer Architecture (ISCA), Virtual.","DOI":"10.1109\/ISCA45697.2020.00047"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Raihan, M.A., Goli, N., and Aamodt, T.M. (2019, January 24\u201326). Modeling Deep Learning Accelerator Enabled GPUs. Proceedings of the 2019 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS), Madison, WI, USA.","DOI":"10.1109\/ISPASS.2019.00016"},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Kandiah, V., Peverelle, S., Khairy, M., Pan, J., Manjunath, A., Rogers, T.G., Aamodt, T.M., and Hardavellas, N. (2021, January 18\u201322). AccelWattch: A Power Modeling Framework for Modern GPUs. Proceedings of the 54th Annual IEEE\/ACM International Symposium on Microarchitecture (MICRO), Virtual.","DOI":"10.1145\/3466752.3480063"},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Fatica, M. (2008, January 24\u201326). CUDA Toolkit and Libraries. Proceedings of the 2008 IEEE Hot Chips 20 Symposium (HCS), Stanford, CA, USA.","DOI":"10.1109\/HOTCHIPS.2008.7476520"},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Gale, T., Zaharia, M., Young, C., and Elsen, E. (2020, January 9\u201319). Sparse GPU Kernels for Deep Learning. Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC), Virtual.","DOI":"10.1109\/SC41405.2020.00021"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Zhu, M., Zhang, T., Gu, Z., and Xie, Y. (2019, January 12\u201316). Sparse Tensor Core: Algorithm and Hardware Co-Design for Vector-Wise Sparse Neural Networks on Modern GPUs. Proceedings of the 52nd Annual IEEE\/ACM International Symposium on Microarchitecture (MICRO), Columbus, OH, USA.","DOI":"10.1145\/3352460.3358269"},{"key":"ref_31","unstructured":"Merity, S., Xiong, C., Bradbury, J., and Socher, R. (2016). Pointer sentinel mixture models. arXiv."},{"key":"ref_32","unstructured":"NVIDIA (2021). NVIDIA AMPERE GA102 GPU Architecture, NVIDIA. Technical Report."}],"container-title":["Future Internet"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1999-5903\/18\/2\/88\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,6]],"date-time":"2026-02-06T16:40:51Z","timestamp":1770396051000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1999-5903\/18\/2\/88"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,6]]},"references-count":32,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2026,2]]}},"alternative-id":["fi18020088"],"URL":"https:\/\/doi.org\/10.3390\/fi18020088","relation":{},"ISSN":["1999-5903"],"issn-type":[{"value":"1999-5903","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,6]]}}}