inference/research
verifiedhigh · 0.88updated 2026-10-03

Stateful Large Language Model Serving with Pensieve (Shi et al., Microsoft)

paper

Fields

id
sources/qlab-pensieve-paper
URL
https://arxiv.org/abs/2312.05516
Publisher
arXiv
Kind
paper
Published
2023-12-09
Accessed
2026-10-03
notes
v3. Multi-tier GPU/CPU conversation-state cache tier. Storage tiering, not quantization - recorded to mark the boundary.

Notes

v3. Multi-tier GPU/CPU conversation-state cache tier. Storage tiering, not quantization - recorded to mark the boundary.

Cross-references

This record references no other records.