Random Access Memory

Random Access Memory (RAM) is a fundamental component of computing systems, and current research focuses on improving its performance, energy efficiency, and reliability for various applications. Key areas of investigation include optimizing DRAM access for deep neural networks (DNNs) through novel data mapping policies and hardware co-design, mitigating errors in DRAM through adaptive methods and fault-tolerant architectures, and exploring processing-in-DRAM (PIM) approaches to accelerate computations. These advancements are crucial for enabling energy-efficient AI at the edge, high-performance computing in data centers, and the development of more robust and reliable computing systems.