DRAM Access Energy

DRAM access energy is a critical bottleneck in modern computing, particularly for energy-intensive applications like deep learning and machine intelligence. Research focuses on reducing this energy consumption through techniques such as developing novel memory architectures (e.g., computational RAM), optimizing data access patterns in neural networks (e.g., through model-agnostic error correction or efficient mapping strategies), and designing hardware accelerators that minimize data movement between memory and processing units. These efforts aim to improve the energy efficiency of computing systems, enabling more powerful and sustainable applications in various fields.

Papers