未登录CD-LLM: A Heterogeneous Multi-FPGA System for Batched Decoding of 70B+LLMs Using a Compute-Dedicated ArchitectureCD-LLM:一种用于批处理解码70B+LLM的计算专用异构多FPGA系统
Ma, Wenheng; Yang, Xinhao; Zeng, Shulin; Liu, Tengxuan; Shen, Libo; Wang, Hongyi; Li, Shiyao; Hong, Ke; Zhu, Zhenhua; Ning, Xuefei; Ho, Tsung-Yi; Dai, Guohao; Wang, Yu
分享
收藏
分享
收藏
分享
收藏