MIDAS: An Energy-Efficient Microscaling Quantization-Based Digital CIM Accelerator With Spatio-Temporal Cyclic Alignment for Generative AI
Abstract
Digital compute-in-memory (DCIM) and the microscaling (MX) data format have emerged as promising hardware and algorithmic solutions, respectively, for energy-efficient generative AI (GenAI) processing. However, their combination introduces challenges, including redundant computation from mantissa alignment, excessive exponent handling overhead, and accuracy loss from static approximation schemes that fail to accommodate block-wise dynamic bit-significance. In this article, we present MIDAS, an MX-optimized DCIM-based GenAI accelerator with three key features: 1) spatio-temporal cyclic alignment (STCA) with sign-magnitude-based assembled column addition (SACA) enables compact weight mapping, high input reusability, and low-latency computation, achieving <inline-formula> <tex-math notation="LaTeX">$1.72\times $ </tex-math></inline-formula> higher macro energy efficiency; 2) MX-aware minimum exponent post-detection (MEPD) eliminates redundant maximum exponent detection by exploiting the inherent characteristics of MX quantization, reducing exponent handling power, area, and latency by 68.8%, 63.2%, and 22.5%, respectively; and 3) sense amplifier-based dynamic bit (SADB)-significance handling with shared scale-aware channel regrouping (SSCR) dynamically adjusts approximation bit-positions according to each block’s shared scale, improving energy efficiency by 27.3% while minimizing accuracy loss. MIDAS is fabricated in Samsung 28-nm CMOS technology and occupies a 7.56 mm<sup>2</sup> die area. MIDAS achieves macro energy and area efficiencies of 96.3 TFLOPS/W and 0.69 TFLOPS/mm<sup>2</sup>, respectively, under MXFP4. Compared to state-of-the-art DCIMs, MIDAS demonstrates <inline-formula> <tex-math notation="LaTeX">$1.58\times $ </tex-math></inline-formula> higher system energy efficiency and <inline-formula> <tex-math notation="LaTeX">$3.32\times $ </tex-math></inline-formula> higher macro energy efficiency, validating its effectiveness as an MX-optimized AI accelerator for next-generation GenAI applications.
