QPART chooses, for each edge inference request, a layer split point and per-layer bit widths that minimize time, energy, and server cost subject to an accuracy budget, cutting communication payload by over 80% with measured accuracy loss below 1%.
Edge computing: Vision and challenges,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
QPART: Adaptive Model Quantization and Dynamic Workload Balancing for Accuracy-aware Edge Inference
QPART chooses, for each edge inference request, a layer split point and per-layer bit widths that minimize time, energy, and server cost subject to an accuracy budget, cutting communication payload by over 80% with measured accuracy loss below 1%.