First Implementation of Monolithic Integrated Cim with 1Mb Ultra-High-Density 8-Layer 3D VRRAM, Achieving High Computing Density ($204.8 \text{GOPs} / \text{mm}^{2}$) and FoM ($\mathbf{2. 1 3} \times \mathbf{1 0}^{\mathbf{6}} \text{GOPS}^{\mathbf{2}}

This work firstly implemented the Monolithic Integration of ultra-high density 1 Mb 8-layer 3D Vertical RRAM based Computing-In-Memory chip (MI-3DCIM) for efficient scientific computing. The 3D VRRAM performs in-situ matrix-vector multiplication (MVM) and transmits data to underlying computational circuits via the BEOL metal interconnects, which is embedded with the on-chip processor, enabling a fully monolithic SoC. Benefiting from the reduced latency and energy consumption, the MI-3DCIM exhibits ultra-high computational density (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$204.8 \text{GOPS} / \text{mm}^{2}$</tex>), high energy efficiency (10.4 TOPS/W) and superior Figure of Merit (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$2.13 \times 10^{6} \text{GOPS}^{2} / \mathrm{W} / \text{mm}^{2}$</tex>). This novel 3DCIM chip greatly enhances the capability for complex tasks.

Authors

Institutions

Publication Details

Published
2025-06-08
DOI
https://doi.org/10.23919/vlsitechnologyandcir65189.2025.11075209
Citations
1
Primary Topic
Semiconductor materials and devices
Type
article
Field-Weighted Citation Impact
0.73

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

First Implementation of Monolithic Integrated Cim with 1Mb Ultra-High-Density 8-Layer 3D VRRAM, Achieving High Computing Density ($204.8 \text{GOPs} / \text{mm}^{2}$) and FoM ($\mathbf{2. 1 3} \times \mathbf{1 0}^{\mathbf{6}} \text{GOPS}^{\mathbf{2}}

Yu Liu, Xiaoxin Xu, Shuaidi Zhang, Zhen Fan et al.
1 citations
Semiconductor materials and devices
0.73
article

First Implementation of Monolithic Integrated Cim with 1Mb Ultra-High-Density 8-Layer 3D VRRAM, Achieving High Computing Density ($204.8 \text{GOPs} / \text{mm}^{2}$) and FoM ($\mathbf{2. 1 3} \times \mathbf{1 0}^{\mathbf{6}} \text{GOPS}^{\mathbf{2}}

Yu Liu, Xiaoxin Xu, Shuaidi Zhang, Zhen Fan, Peiwen Zhang, Feng Zhang, Cailian Ma, Kaiping Zhang, Shengjie Zhao, Yue Wu, Long Nie, Weiye Tang, Wenxuan Sun, Yao Li, Ming Liu
article en
1 citations

Abstract

This work firstly implemented the Monolithic Integration of ultra-high density 1 Mb 8-layer 3D Vertical RRAM based Computing-In-Memory chip (MI-3DCIM) for efficient scientific computing. The 3D VRRAM performs in-situ matrix-vector multiplication (MVM) and transmits data to underlying computational circuits via the BEOL metal interconnects, which is embedded with the on-chip processor, enabling a fully monolithic SoC. Benefiting from the reduced latency and energy consumption, the MI-3DCIM exhibits ultra-high computational density (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$204.8 \text{GOPS} / \text{mm}^{2}$</tex>), high energy efficiency (10.4 TOPS/W) and superior Figure of Merit (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$2.13 \times 10^{6} \text{GOPS}^{2} / \mathrm{W} / \text{mm}^{2}$</tex>). This novel 3DCIM chip greatly enhances the capability for complex tasks.

Beijing Institute of Technology (CN), Chinese Academy of Sciences (CN), Institute of Microelectronics (CN)
National Natural Science Foundation of China, National Key Research and Development Program of China
Industry, innovation and infrastructure
Openalex Percentile: Top 25%
Semiconductor materials and devices
0.73
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.