An I/O Characterizing Study of Offloading LLM Models and KV Caches to NVMe SSD
Metadata
- Authors: Zebin Ren, Krijn Doekemeijer, Tiziano De Matteis, Christian Pinto, Radu Stoica, Animesh Trivedi
- Type: Workshop paper
- Publish date: 2025-03-30
Availability
- Paper at AtLarge: 2025-cheops-llm.pdf
- ACM: https://dl.acm.org/doi/pdf/10.1145/3719330.3721230
- Code: https://github.com/stonet-research/cheops25-IO-characterization-of-LLM-model-kv-cache-offloading-nvme