Oneiros: KV Cache Optimization through Parameter Remapping for Multi-tenant LLM Serving
Open Access
- 19 November 2025
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 88-101
Abstract
No abstract availableAll Related Versions
Funding Information
- NSF (National Science Foundation) (2326894, 2425655)
This publication has 24 references indexed in Scilit:
- ParvaGPU: Efficient Spatial GPU Sharing for Large-Scale DNN Inference in Cloud Environments2024
- A Mess of Memory System Benchmarking, Simulation and Application Profiling2024
- A survey on LLM-based multi-agent systems: workflow, infrastructure, and challengesVicinagearth, 2024
- Splitwise: Efficient Generative LLM Inference Using Phase Splitting2024
- iGniter: Interference-Aware GPU Resource Provisioning for Predictable DNN Inference in the CloudIEEE Transactions on Parallel and Distributed Systems, 2022
- ZeRO: Memory optimizations Toward Training Trillion Parameter Models2020
- Leveraging Power-Performance Relationship of Energy-Efficient Modern DRAM DevicesIEEE Access, 2018
- Resource Central2017
- GPU Computing Pipeline Inefficiencies and Optimization Opportunities in Heterogeneous CPU-GPU Processors2015
- Few-to-ManyACM SIGPLAN Notices, 2015