PipeMax integrates pipeline parallelism with offloading to achieve up to 2.51x higher throughput than vLLM for offline LLM inference on commodity 8-GPU servers.
arXiv preprint arXiv:2503.01328 , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
PipeMax: Enhancing Offline LLM Inference on Commodity GPU Servers
PipeMax integrates pipeline parallelism with offloading to achieve up to 2.51x higher throughput than vLLM for offline LLM inference on commodity 8-GPU servers.