Processing math: 100%
IGZO CIM: Enabling In-Memory Computations Using Multilevel Capacitorless Indium–Gallium–Zinc–Oxide-Based Embedded DRAM Technology | IEEE Journals & Magazine | IEEE Xplore

IGZO CIM: Enabling In-Memory Computations Using Multilevel Capacitorless Indium–Gallium–Zinc–Oxide-Based Embedded DRAM Technology


Abstract:

Compute-in-memory (CIM) is a promising approach for efficiently performing data-centric computing (such as neural network computations). Among the multiple semiconductor ...Show More

Abstract:

Compute-in-memory (CIM) is a promising approach for efficiently performing data-centric computing (such as neural network computations). Among the multiple semiconductor memory technologies, embedded DRAM (eDRAM), which integrates the DRAM bit cell with high-performance logic transistors, can enable efficient CIM designs. However, the silicon-based eDRAM technology suffers from poor retention time-incurring significant refresh power overhead. However, eDRAM using back-end-of-line (BEOL) integrated C -axis aligned crystalline (CAAC) indium–gallium–zinc–oxide (IGZO) transistors, exhibiting extreme low leakage, is a promising memory technology with lower refresh power overhead. A long retention time in IGZO eDRAM can enable multilevel cell functionality, which can improve its efficacy in CIM applications. In this article, we explore a capacitorless IGZO eDRAM-based multilevel cell, capable of storing 1.5 bits/cell for CIM designs focused on deep neural network (DNN) inference applications. We perform a detailed design space exploration of IGZO eDRAM sensitivity to process temperature variations for read, write, and retention operations followed by architecture-level simulations comparing performance and energy for different workloads. The effectiveness of IGZO eDRAM-based CIM architecture is evaluated using a representative neural network, and the proposed approach achieves 82% Top-1 inference accuracy for the CIFAR-10 dataset, compared with 87% software accuracy with high bit cell storage density.
Page(s): 35 - 43
Date of Publication: 04 July 2022
Electronic ISSN: 2329-9231

Funding Agency:

References is not available for this document.

Select All
1.
Intel Xeon Platinum 8380 HL Processor, Intel, Santa Clara, CA, USA, 1999.
2.
NVIDIA Tesla V100 GPU Architecture, NVIDIA.
3.
Z. Jia, B. Tillman, M. Maggioni, and D. P. Scarpazza, “Dissecting the graphcore IPU architecture via microbenchmarking,” 2019, arXiv:1912.03413.
4.
S. K. Gonugondla, M. Kang, and N. Shanbhag, “A 42 pJ/decision 3.12TOPS/W robust in-memory machine learning classifier with on-chip training,” in IEEE Int. Solid-State Circuits Conf. (ISSCC) Dig. Tech. Papers, Feb. 2018, pp. 490–492.
5.
X. Si, “24.5 A twin-8T SRAM computation-in-memory macro for multiple-bit CNN-based machine learning,” in IEEE Int. Solid-State Circuits Conf. (ISSCC) Dig. Tech. Papers, Feb. 2019, pp. 396–398.
6.
B. Zhang, “90% yield production of polymer nano-memristor for in-memory computing,” Nature Commun., vol. 12, no. 1, pp. 1–11, Dec. 2021.
7.
S. Sagar, K. U. Mohanan, S. Cho, L. A. Majewski, and B. C. Das, “Emulation of synaptic functions with low voltage organic memtransistor for hardware oriented neuromorphic computing,” Sci. Rep., vol. 12, no. 1, pp. 1–12, Dec. 2022.
8.
S. Mittal, J. S. Vetter, and D. Li, “A survey of architectural approaches for managing embedded DRAM and non-volatile on-chip caches,” IEEE Trans. Parallel Distrib. Syst., vol. 26, no. 6, pp. 1524–1537, Jun. 2015.
9.
G. Fredeman, “17.4 A 14 nm 1.1 Mb embedded DRAM macro with 1 ns access,” in IEEE Int. Solid-State Circuits Conf. (ISSCC) Dig. Tech. Papers, Feb. 2015, pp. 1–3.
10.
N. Kurd, “Haswell: A family of IA 22 nm processors,” IEEE J. Solid-State Circuits, vol. 50, no. 1, pp. 49–58, Jan. 2015.
11.
M. Oota, “3D-stacked CAAC-In-Ga-Zn oxide FETs with gate length of 72 nm,” in IEDM Tech. Dig., Dec. 2019, pp. 2–3.
12.
S. R. S. Raman, S. Xie, and J. P. Kulkarni, “Compute-in-eDRAM with backend integrated indium gallium zinc oxide transistors,” in Proc. IEEE Int. Symp. Circuits Syst. (ISCAS), May 2021, pp. 1–5.
13.
A. Belmonte, “Capacitor-less, long-retention (> 400s) DRAM cell paving the way towards low-power and high-density monolithic 3D DRAM,” in IEDM Tech. Dig., Dec. 2020, pp. 2–28.
14.
A. Belmonte, “Tailoring IGZO-TFT architecture for capacitorless DRAM, demonstrating >10³s retention, >10¹¹ cycles endurance and L G scalability down to 14 nm,” in IEDM Tech. Dig., Dec. 2021, pp. 6–10.
15.
Y. S. Chauhan, FinFET Modeling for IC Simulation and Design: Using the BSIM-CMG Standard. New York, NY, USA : Academic, 2015.
16.
T. Atsumi, “DRAM using crystalline oxide semiconductor for access transistors and not requiring refresh for more than ten days,” in Proc. 4th IEEE Int. Memory Workshop, May 2012, pp. 1–4.
17.
A. Belmonte, “Capacitor-less, long-retention (>400s) DRAM cell paving the way towards low-power and high-density monolithic 3D DRAM,” in IEDM Tech. Dig., Dec. 2020, p. 28.
18.
C. Hu, “BSIM-IMG: A turnkey compact model for fully depleted technologies,” in Proc. IEEE Int. SOI Conf. (SOI), Oct. 2012, pp. 1–24.
19.
Y. Kim, W. Yan, and O. Mutlu, “Ramulator: A fast and extensible DRAM simulator,” IEEE Comput. Archit. Lett. vol. 15, no. 1, pp. 45–49, Jan. 2016.
20.
512Mb DDR2 SDRAM Component Data Sheet: MT47H128M4B6-25, MICRON.
21.
J. Gómez-Luna, I. E. Hajj, I. Fernandez, C. Giannoula, G. F. Oliveira, and O. Mutlu, “Benchmarking a new paradigm: Experimental analysis and characterization of a real processing-in-memory system,” IEEE Access, vol. 10, pp. 52565–52608, 2022, doi: 10.1109/ACCESS.2022.3174101.
22.
B. C. Lee, E. Ipek, O. Mutlu, and D. Burger, “Architecting phase change memory as a scalable dram alternative,” in Proc. 36th Annu. Int. Symp. Comput. Archit. (ISCA), 2009, pp. 2–13.
23.
X. Yang, “An in-memory-computing charge-domain ternary CNN classifier,” in Proc. IEEE Custom Integr. Circuits Conf. (CICC), Apr. 2021, pp. 1–2.

References

References is not available for this document.