LUT-DLA: Lookup Table As Efficient Extreme Low-Bit Deep Learning Accelerator
International Symposium on High-Performance Computer Architecture（2025）
关键词
Lookup Table,Neural Network,Space Exploration,Power Efficiency,Manhattan Distance,Numerical Representation,Accuracy Drop,Hardware Accelerators,Vector Quantization,Bit-width,Hardware Parameters,Accuracy Of Model,Similarity Measure,Power Consumption,Search Space,Input Vector,Matrix Multiplication,Vector Of Length,Identification Performance,Design Space,Hardware Architecture,Design Space Exploration,Approximate Computation,Hardware Cost,Similarity Calculation,On-chip Memory,Improvement In Power,Backward Pass,Input Matrix,Representation Of The Input Data
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要