Poster Type: Research Posters
Author: Mona Moghadampanah (Virginia Tech), Xiangche Li (Virginia Tech), Farhana Amin (Virginia Tech), Dimitrios S. Nikolopoulos (Virginia Tech)
Abstract: Existing diffusion LLM (dLLM) serving systems focus on optimizing latency and throughput, while their energy efficiency remains less studied. We show that dLLM decoding moves between compute-bound cache-refresh and memory-bound cache-reuse phases, and that decoding configurations create a broad accuracy–energy trade-off. We present E-dLLM, which combines configuration selection with phase-aware dynamic voltage and frequency scaling (DVFS). On LLaDA-8B, reactive DVFS reduces energy by 17% with less than 1% latency overhead, while preserving model outputs.
Best Poster Finalist (BP): no
Poster: PDF
Poster Summary: PDF