ARMay 29

A Reconfigurable Computing In-Memory Macro with Charge-sharing-based Weighted Accumulator

arXiv:2605.3081482.4h-index: 16
AI Analysis

This work offers significant improvements in efficiency and performance for researchers and engineers developing analog computing-in-memory systems, particularly in reducing ADC overhead and latency.

This paper addresses challenges in SRAM-based analog computing-in-memory by proposing a reconfigurable 256x128 array. It achieves a 9x improvement in IMADC area overhead, reduces latency by up to 6.6x, and improves read bitline voltage by 3.5x.

SRAM-based analog computing-in-memory demonstrates outstanding efficiency. However, it faces three critical challenges: significant ADC overhead, high latency for multi-bit inputs, and limited read bitline voltage. To address these issues, this work proposes a multi-bit highly reconfigurable 256x128 in-memory computing array supporting 1-7b input, 2-4b weight, and 1-7b output. Three key innovations are introduced: 1) The IMADC occupies only 3% area overhead, achieving a 9x improvement compared to previous IMADC; 2) The BSCHA reduces latency by 1.9x and 6.6x compared to traditional pulse-width modulation (PWM) and bit-slicing modes, respectively; 3) A dual-8T bitcell enabling ternary weight storage through a decoupled read path, integrated with a read wordline under-driven cascode technique, improves linearity of unit discharge current by 7x and increases the usable read bitline voltage by 3.5x.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes