A dispatch-aware roofline simulator picks the best LLM serving strategy by predicted goodput without running costly GPU benchmarks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
BestServe: Serving Strategies with Optimal Goodput in Collocation and Disaggregation Architectures
A dispatch-aware roofline simulator picks the best LLM serving strategy by predicted goodput without running costly GPU benchmarks.