KiHyun Kim
MS/PhD Integrated student
Department of Computer Science and Engineering Sogang University , Seoul, South Korea
Member of Data-Intensive Computing & AI Systems Laboratory (DISCOS)
Contacts: kion777@sogang.ac.kr, zexyprofessional16@gmail.com
Research Interests
LLM Inference System
KV cache Management
Cost-Efficient Computing
In-Storage Processing, Computational Storage Device
Database System
Operating System
Education
MS/PhD Intergrated Program in Computer Science and Engineering, Sogang University, Seoul South Korea 2022 - Current
Bachelor Degree of Computer Science & Engineering, Sogang University – Seoul, South Korea 2016 - 2022
2026
K. Kim, S. Jang, M. Heo, J. Kim, J. J. Kim, and Y. Kim, “AegisKV: Runtime Headroom Control for Online LLM Serving via Adaptive Attention Offloading.”
In Proceedings of the ACM International Conference on Information and Knowledge Management (CIKM), 2026.
S. Jang, K. Kim, J. Kim, M. Heo, D. Li, and Y. Kim, “PAVE: Partial KV Cache Eviction for Fine-Grained KV Reclamation in LLM Serving.”
In Proceedings of the IEEE International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems (MASCOTS), 2026.
J. Kim, K. Kim, H. Chung, J. Yang, J. J. Kim, D. Li, and Y. Kim,
“LLM-Pilot: SLO-Aware and Cost-Efficient LLM Serving on Public Cloud VM Clusters via Offloading.”
In Proceedings of the IEEE International Symposium on Cluster, Cloud and Internet Computing (CCGRID), 2026.
2025
K. Kim, J. Kim, J. J. Kim, D. Li, and Y. Kim,
“FLEXLLM: Flexible and Cost-Efficient LLM Serving with Heterogeneous GPUs.”
In Proceedings of the IEEE Int’l Symp. Modeling, Analysis and Simulation of Computer and Telecommunication Systems (MASCOTS), 2025.
K. Kim, J. Kim, H. Chung, M.-H. Cha, H.-Y. Kim, and Y. Kim,
“Cost-Efficient LLM Serving in the Cloud: VM Selection with KV Cache Offloading.”
In Proceedings of the IEEE International Conference on Cloud Computing (CLOUD), 2025.
H. Lee, K. Kim, J. Kim, J. So, M.-H. Cha, H.-Y. Kim, J. J. Kim, and Y. Kim,
“Shared Disk KV Cache Management for Efficient Multi-Instance Inference in RAG-Powered LLMs.”
In Proceedings of the IEEE International Conference on Cloud Computing (CLOUD), 2025.
2024
Y. Kim, K. Kim, Y. Cho, J. Kim, A. Khan, K.-D. Kang, B.-S. An, M.-H. Cha, H.-Y. Kim, and Y. Kim,
“DeepVM: Integrating Spot and On-Demand VMs for Cost-Efficient Deep Learning Clusters in the Cloud.”
Proceedings of the IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID), 2024.
Y. Cho, Y. Kim, K. Kim, J. Kim, H.-Y. Kim, and Y. Kim,
“Optimizing Multi-level Checkpointing for Distributed Deep Learning Workloads on Cloud Spot VM Clusters.”
IEEE Access, vol. 12, pp. 116891–116904, Aug. 2024.
2023
D. Min, K. Kim, C. Moon, A. Khan, S. Lee, C. Youn, W. S. Chung, and Y. Kim,
“A Multi-Tenant Key-Value SSD with Secondary Index for Search Query Processing and Analysis.”
ACM Transactions on Embedded Computing Systems (TECS), vol. 22, no. 4, Article 65, July 2023.
2022
S. Jamil, A. Khan, K. Kim, and Y. Kim,
“DENKV: Addressing Design Trade-offs of Key-Value Stores for Scientific Applications.”
Proceedings of the 7th International Parallel Data Systems Workshop (PDSW), co-located with SC22, Dallas, TX, Nov. 2022.
K. Kim, J. Kim, and Y. Kim,
"Accelerating RAG-Based LLM Inference Using DRAM-Based Storage."
32nd Korean Conference on Semiconductors (KCS), High1 Resort, Korea, Feb. 2025.
J. Kim, K. Kim, Y. Kim, M.-H. Cha, H.-Y. Kim, J. J. Kim, and Y. Kim,
"Resolving I/O Bottlenecks through Partition Prediction and Preloading in Large-Scale Disk-based GNN Training."
Korea Software Congress (KSC 2024), Yeosu Expo Convention Center, Korea, Dec. 2024.
H. Lee, K. Kim, J. So, M.-H. Cha, H.-Y. Kim, and Y. Kim,
"High-Performance LLM RAG System Utilizing Disk-Based KV Cache in Vector Database."
Korea Software Congress (KSC 2024), Yeosu Expo Convention Center, Korea, Dec. 2024.
K. Kim, Y. Noh, J. Park, E. Kim, and Y. Kim,
"Detecting Fraudulent Reward Acquisition in Blockchain-Based Reward Applications."
Korea Computer Congress (KCC 2023), Ramada Plaza Jeju, Korea, June 2023.
K. Kim and Y. Kim,
"Host-Coordinated Deduplication for Key-Value SSDs: A Feasibility Study."
30th Korean Conference on Semiconductors (KCS), High1 Resort, Korea, Feb. 2023.
Work Experience
Distributed Computing & Operating Systems Laboratory (DISCOS) 03/2021 - 08/2021
Responsibilities:
Implement LRU based Bloom-Filter Cache in iLSM-KVSSD platform.
Prepare data and information for making experimental report
· Implement LRU based Bloom-Filter Cache in iLSM- KVSSD Platform.
· Prepare data and information for making experimental report.
US Department of Energy's Oak Ridge National Laboratory, USA. 07/2022 - 08/2022
Projects
A study on Data Deduplication of Parallel File Systems for High Performance Computing, Korea Institute of Science and Technology Information (KISTI) 05/2022 - current
OpenCV Project : "Thimble" : Fingerprint blurring Application, Capstone Design 09/2021 - 12/2021
Develop fingerprint blurring mechanism on image and streaming video using media-pipe library.
Arduino Project: Smart Water Bottle and IOS Application, Capstone Design 2 09/2021 - 12/2021
Develop Smart Water Bottle which tracks user’s intake and gives feedback to user.
Develop IOS data-transfer mechanism using react- native Bluetooth API.