3 papers
cs.AR2026
Rethinking LLM Inference Bottlenecks: Insights from Latent Attention and Mixture-of-Experts
Sungmin Yun, Seonyong Park, Hwayong Nam +10
Computational workloads composing traditional transformer models are starkly bifurcated. Multi-Head Attention (MHA) and Grouped-Query Attention are memory-bound due to low arithmet…
cs.AR2025
SSD Offloading for LLM Mixture-of-Experts Weights Considered Harmful in Energy Efficiency
Kwanhee Kyung, Sungmin Yun, Jung Ho Ahn
Large Language Models (LLMs) applying Mixture-of-Experts (MoE) scale to trillions of parameters but require vast memory, motivating a line of research to offload expert weights fro…
cs.AR2024
Duplex: A Device for Large Language Models with Mixture of Experts, Grouped Query Attention, and Continuous Batching
Sungmin Yun, Kwanhee Kyung, Juhwan Cho +6
Large language models (LLMs) have emerged due to their capability to generate high-quality content across diverse contexts. To reduce their explosively increasing demands for compu…