Endor: Exploit Nearly-Decode-Only Opportunities of LLM Reasoning on Near-Memory Architecture.
Jun Liu, Tianlang Zhao, Shiyi Liu, Jiancai Ye, Lin Li, Zhen Yu, Li Ding, Hao Zhou, Zhenhua Zhu, Xuefei Ning, Yuan Xie, Yu Wang, Guohao Dai
Browse the full DATE paper archive.