TY - GEN
T1 - Power and Thermal Modeling of In-3D-Memory Computing
AU - Han, Jun Han
AU - West, Robert E.
AU - Torres-Castro, Karina
AU - Swami, Nathan
AU - Khan, Samira
AU - Stan, Mircea
N1 - Publisher Copyright:
© 2021 IEEE.
PY - 2021/3/3
Y1 - 2021/3/3
N2 - In-memory computing is an efficient big-data processing method that allows for performance scalability. Although 3D-memory is an established technology to integrate in-memory computing due to the 3D-stacked logic and memory layers, there are thermal issues and power budgets that restrict the integration of this concept in a 3D-stacked memory. In this paper, we investigate the power and thermal behavior of the in-3D-memory computing device based on thermal circuit modeling. When the logic layers consume 1 W/cm2 in each layer, the maximum power available in the logic layer is 11.4 W/cm2 and 7.3 W/cm2 for the 4 and 16-layer stack. For high-rise stacked memory, passive heat sinks cannot sustain the operating temperature below 85 C. In this work, the validity of the proposed simulation results is compared with a conventional simulator.
AB - In-memory computing is an efficient big-data processing method that allows for performance scalability. Although 3D-memory is an established technology to integrate in-memory computing due to the 3D-stacked logic and memory layers, there are thermal issues and power budgets that restrict the integration of this concept in a 3D-stacked memory. In this paper, we investigate the power and thermal behavior of the in-3D-memory computing device based on thermal circuit modeling. When the logic layers consume 1 W/cm2 in each layer, the maximum power available in the logic layer is 11.4 W/cm2 and 7.3 W/cm2 for the 4 and 16-layer stack. For high-rise stacked memory, passive heat sinks cannot sustain the operating temperature below 85 C. In this work, the validity of the proposed simulation results is compared with a conventional simulator.
KW - 3D-IC
KW - 3D-memory
KW - In-memory computing
KW - power simulation
KW - Processing-in-memory
KW - Thermal modeling
UR - https://www.scopus.com/pages/publications/85104652605
U2 - 10.1109/ISDCS52006.2021.9397913
DO - 10.1109/ISDCS52006.2021.9397913
M3 - Conference contribution
AN - SCOPUS:85104652605
T3 - 4th International Symposium on Devices, Circuits and Systems, ISDCS 2021 - Conference Proceedings
BT - 4th International Symposium on Devices, Circuits and Systems, ISDCS 2021 - Conference Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 4th International Symposium on Devices, Circuits and Systems, ISDCS 2021
Y2 - 3 March 2021 through 5 March 2021
ER -