TY - JOUR
T1 - Improved MEOL and BEOL Parasitic-Aware Design Technology Co-Optimization for 3 nm Gate-All-Around Nanosheet Transistor
AU - Sun, Yabin
AU - Wang, Meng
AU - Li, Xianglong
AU - Hu, Shaojian
AU - Liu, Ziyu
AU - Liu, Yun
AU - Li, Xiaojin
AU - Shi, Yanling
N1 - Publisher Copyright:
© 1963-2012 IEEE.
PY - 2022/2/1
Y1 - 2022/2/1
N2 - In this article, an improved parasitic-aware design technology co-optimization (DTCO) for gate-all-around nanosheet field effect transistor (GAA-NSFET) at 3 nm node is proposed. The presented DTCO flow owns two distinct features. First, a novel de-embedding strategy is designed to avoid the repeated calculation of gate-source/drain contact capacitance. Second, the parasitic resistance of the middle-end-of-line (MEOL) and back-end-of-line (BEOL) is accurately extracted, combing the front-end-of-line (FEOL) simulation and the calculation of MEOL/BEOL equivalent interconnect length. The power, performance, and area (PPA) of the benchmark circuit [15-stage ring oscillator (RO)] are collaboratively optimized. Considering the limitation of contacted gate pitch (CGP) and the process effects, the compromise of structure parameters is studied. GAA-NSFET architecture with 48% reduction in power consumption, 26% increase in speed, and 46% reduction in area is achieved, satisfying the scaling requirement from 5 to 3 nm node. All data here provide an optimization and design foundation for GAA-NSFET in future 3 nm technology node.
AB - In this article, an improved parasitic-aware design technology co-optimization (DTCO) for gate-all-around nanosheet field effect transistor (GAA-NSFET) at 3 nm node is proposed. The presented DTCO flow owns two distinct features. First, a novel de-embedding strategy is designed to avoid the repeated calculation of gate-source/drain contact capacitance. Second, the parasitic resistance of the middle-end-of-line (MEOL) and back-end-of-line (BEOL) is accurately extracted, combing the front-end-of-line (FEOL) simulation and the calculation of MEOL/BEOL equivalent interconnect length. The power, performance, and area (PPA) of the benchmark circuit [15-stage ring oscillator (RO)] are collaboratively optimized. Considering the limitation of contacted gate pitch (CGP) and the process effects, the compromise of structure parameters is studied. GAA-NSFET architecture with 48% reduction in power consumption, 26% increase in speed, and 46% reduction in area is achieved, satisfying the scaling requirement from 5 to 3 nm node. All data here provide an optimization and design foundation for GAA-NSFET in future 3 nm technology node.
KW - Back-end-of-line (BEOL)
KW - compact model (CM)
KW - design technology co-optimization (DTCO)
KW - gate-all-around nanosheet field effect transistor (GAA-NSFET)
KW - middle-end-of-line (MEOL)
KW - parasitic extraction
UR - https://www.scopus.com/pages/publications/85122300220
U2 - 10.1109/TED.2021.3135247
DO - 10.1109/TED.2021.3135247
M3 - 文章
AN - SCOPUS:85122300220
SN - 0018-9383
VL - 69
SP - 462
EP - 468
JO - IEEE Transactions on Electron Devices
JF - IEEE Transactions on Electron Devices
IS - 2
ER -