Development and Opportunity of Advanced Packaging Technology

doi:10.3981/j.issn.2097-0781.2022.03.009

Science and Technology Foresight ›› 2022, Vol. 1 ›› Issue (3): 101-114.DOI: 10.3981/j.issn.2097-0781.2022.03.009

• Review and Commentary • Previous Articles Next Articles

Development and Opportunity of Advanced Packaging Technology

CAO Liqiang(), HOU Fengze(), WANG Qidong, LIU Fengman, LI Jun, DING Fei, SUN Siwei, ZHOU Yunyan

Institute of Microelectronics of the Chinese Academy of Sciences, Beijing 100029, China

Received:2022-07-31 Revised:2022-08-24 Online:2022-09-20 Published:2022-11-04
Contact: HOU Fengze

先进封装技术的发展与机遇

曹立强(), 侯峰泽(), 王启东, 刘丰满, 李君, 丁飞, 孙思维, 周云燕

中国科学院微电子研究所,北京 100029

通讯作者: 侯峰泽
作者简介:曹立强,研究员,博士研究生导师。现任中国科学院微电子研究所副所长,国家科技重大专项02专项总体组专家。主要研究方向为系统级封装及三维集成。作为项目/课题负责人,承担多项国家科技重大专项、国家创新团队国际合作伙伴计划、国家“973”计划、国家自然科学基金项目。合作撰写专著6部。发表学术论文200余篇。申请发明专利180余项,获授权发明专利60余项。电子信箱： caoliqiang@ime.ac.cn。
侯峰泽,副研究员。主要研究方向为微电子/功率电子先进封装、芯粒异质集成、热管理和热机械可靠性。作为项目/课题负责人,承担1项国家自然科学基金面上项目、1项国家科技重大专项02专项课题、1项装备预研项目和多项企业联合研发课题。在学科顶级期刊TPEL、ATE、TED、JESTPE和TCPMT等发表学术论文50余篇。获授权发明专利20余项。电子信箱： houfengze@ime.ac.cn。
基金资助:
国家自然科学基金(62174177);国家自然科学基金(U21A20504)

Abstract

Abstract:

Advanced packaging technology, whose drive has evolved from high-end smartphones to high-performance computing and artificial intelligence (AI), mainly involves high-performance processors, memories, and AI training and inference. As the development of integrated circuits is constrained by the memory, area, power consumption, and function, the advanced packaging technology with chiplet heterogeneous integration at its core will be the key path and breakthrough for the development of integrated circuits. This paper reviews some advanced packaging technologies of milestone significance worldwide and presents the development status and advantages of advanced packaging in China. After that, it analyzes the gap between China and the world’s advanced level and put forward some suggestions for the development of China’s advanced packaging technology.

Key words: high-performance computing, artificial intelligence, chiplet, heterogeneous integration, advanced packaging

摘要：

近年来,先进封装技术的内驱力已从高端智能手机领域演变为高性能计算和人工智能等领域,涉及高性能处理器、存储器、人工智能训练和推理等。当前集成电路的发展受“四堵墙”（“存储墙”“面积墙”“功耗墙”和“功能墙”）制约,以芯粒（Chiplet）异质集成为核心的先进封装技术,将成为集成电路发展的关键路径和突破口。文章概述近年来国际上具有“里程碑”意义的先进封装技术,阐述中国大陆先进封装领域发展的现状与优势,分析中国大陆先进封装关键技术与世界先进水平的差距,最后对未来中国大陆先进封装发展提出建议。

关键词: 高性能计算, 人工智能, 芯粒, 异质集成, 先进封装

CAO Liqiang, HOU Fengze, WANG Qidong, LIU Fengman, LI Jun, DING Fei, SUN Siwei, ZHOU Yunyan. Development and Opportunity of Advanced Packaging Technology[J]. Science and Technology Foresight, 2022, 1(3): 101-114.

曹立强, 侯峰泽, 王启东, 刘丰满, 李君, 丁飞, 孙思维, 周云燕. 先进封装技术的发展与机遇[J]. 前瞻科技, 2022, 1(3): 101-114.

Figures/Tables 22

References 27

[1]	Zhang G Q, Graef M, van Roosmalen F. The rationale and paradigm of “more than Moore”[C]// 56th Electronic Components and Technology Conference 2006. Piscataway: IEEE Press, 2006, doi: 10.1109/ECTC.2006.1645639. DOI
[2]	Bishop M D, Wong H S P, Mitra S, et al. Monolithic 3-D integration[J]. IEEE Micro, 2019, 39(6): 16-27. DOI
[3]	Gholami A. AI and memory wall[EB/OL]. (2021-05-29)[2022-06-10]. https://medium.com/riselab/ai-and-memory-wall-2cb4265cb0b8.
[4]	Su L T, Naffziger S, Papermaster M. Multi-chip technologies to unleash computing performance gains over the next decade[C]// IEEE International Electron Devices Meeting (IEDM). Piscataway: IEEE Press, 2018, doi: 10.1109/IEDM.2017.8268306. DOI
[5]	Mahajan R, Penmecha B, Radhakrishman K. Advanced packaging architectures for heterogeneous integration[C]. MechSE Seminars. 190 Engineering Sciences Building, 2021.
[6]	Tummala R R. Moore,s law meets its match (system-on-package)[J]. IEEE Spectrum, 2006, 43(6): 44-49.
[7]	Yeric G. Three dimensions in 3DIC-Part I[EB/OL].(2018-04-02)[2022-08-10]. https://community.arm.com/arm-research/b/articles/posts/three-dimensions-in-3dic-part-1.
[8]	Su L. Cost per yielded mm² for a 250 mm² die[C]// IEEE International Electron Devices Meeting (IEDM). San Francisco:2017.
[9]	Gomes W, Koker A, Stover P, et al. Ponte Vecchio: A multi-tile 3D stacked processor for exascale computing[C]// IEEE International Solid-State Circuits Conference (ISSCC). Piscataway: IEEE Press, 2022, doi: 10.1109/ISSCC42614.2022.9731673. DOI
[10]	Mujtaba H. AMD discloses its multi-layer chiplet design era, starting with Zen 3 with 3D stacked V-cache technology[EB/OL].(2021-08-22)[2022-05-10]. https://wccftech.com/amd-discloses-multi-layer-chip-design-era-starting-with-zen-3-with-3d-stacked-v-cache-technology/.
[11]	Vivet P, Guthmulle E, Thonnart Y, et al. IntAct: A 96-core processor with six chiplets 3D-stacked on an active interposer with distributed interconnects and integrated power management[J]. IEEE Journal of Solid-State Circuits, 2021, 56(1): 79-97. DOI URL
[12]	Hruska J. TSMC announces its first 16 nm FinFET networking chip: 32-core ARM Cortex-A57[EB/OL].(2014-09-26)[2022-05-10]. https://www.extremetech.com/computing/190941-tsmc-announces-its-first-16nm-finfet-net-working-chip-32-core-arm-cortex-a57.
[13]	William S. Semiconductive wafer and method of making the same: US, 3044909[P]. 1962-07-17.
[14]	Huang P K, Lu C Y, Wei W H, et al. Wafer level system integration of the fifth generation CoWoS®-S with high performance Si interposer at 2500 mm²[C]// 2021 IEEE 71st Electronic Components and Technology Conference. Piscataway: IEEE Press, 2021, doi: 10.1109/ECTC32696.2021.00028. DOI
[15]	Chaware R, Nagarajan K, Ramalingam S. Assembly and reliability challenges in 3D integration of 28 nm FPGA die on a large high density 65 nm passive interposer[C]// 2012 IEEE 62nd Electronic Components and Technology Conference. Piscataway: IEEE Press, 2012, doi: 10.1109/ECTC.2012.6248841. DOI
[16]	Banijamali B, Ramalingam S, Liu H, et al. Outstanding and innovative reliability study of 3D TSV interposer and fine pitch solder micro-bumps[C]// 2012 IEEE 62nd Electronic Components and Technology Conference. Piscataway: IEEE Press, 2012, doi: 10.1109/ECTC.2012.6248847. DOI
[17]	Balaban S. NVIDIA A100 GPU-deep learning benchmark estimates[EB/OL]. (2020-05-22)[2022-05-10]. https://lambdalabs.com/blog/nvidia-a100-gpu-deep-learning-bench-marks-and-architectural-overview/.
[18]	Coudrain P, Charbonnier J, Garnier A, et al. Active interposer technology for chiplet-based advanced 3D system architectures[C]// 2019 IEEE 69th Electronic Components and Technology Conference (ECTC). Piscataway: IEEE Press, 2019, doi: 10.1109/ECTC.2019.00092. DOI
[19]	Gomes W, Khushu S, Ingerly D B, et al. 8.1 lakefield and mobility compute: A 3D stacked 10 nm and 22 FFL hybrid processor system in 12×12 mm², 1 mm package-on-package[C]//2020 IEEE International Solid-State Circuits Conference (ISSCC). Piscataway: IEEE Press, 2020, doi: 10.1109/ISSCC19947.2020.9062957. DOI
[20]	Jun H, Cho J, Lee K, et al. HBM (High Bandwidth Memory) DRAM technology and architecture[C]// 2017 IEEE International Memory Workshop (IMW). Piscata-way: IEEE Press, 2017, doi: 10.1109/IMW.2017.7939084. DOI
[21]	Chun S R, Kuo T H, Yi H, et al. InFO_SoW (System-on-Wafer) for high performance computing[C]// 2020 IEEE 70th Electronic Components and Technology Conference (ECTC). Piscataway: IEEE Press, 2020, doi: 10.1109/ECTC32862.2020.00013. DOI
[22]	Tesla AI Day 2021 Review—Part 3: DojoProject. Tesla,s new supercomputer[EB/OL]. (2021-10-11)[2022-05-10]. https://towardsdatascience.com/tesla-ai-day-2021-review-part-3-project-dojo-teslas-new-supercomputer-715d102dbb29.
[23]	Mahajan R, Sane S. Microelectronic package containing silicon patches for high density interconnects: US, 8064.224 B2[P]. 2011-11-22.
[24]	Braunisch H, Aleksov A, Lotz S, et. al. High-speed performance of silicon bridge die-to-die interconnects[C]// 2011 IEEE 20th Conference on Electrical Performance of Electronic Packaging and Systems. Piscataway: IEEE Press, 2011, doi: 10.1109/EPEPS.2011.6100196. DOI
[25]	Starkston D M R, Guzek J S, Chiu C P. Localized high density substrate routing: US, 9136236, B2[P]. 2015-09-15.
[26]	Mahajan R, Sankman R, Patel N, et al. Embedded multi-die interconnect bridge (EMIB): A high density, high bandwidth packaging interconnect[C]// 2016 IEEE 66th Electronic Components and Technology Conference (ECTC). Piscataway: IEEE Press, 2016, doi: 10.1109/ECTC.2016.201. DOI
[27]	Mahajan R, Qian Z, Viswanath R S, et al. Embedded multidie interconnect bridge: A localized, high-density multichip packaging interconnect[J]. IEEE Transactions on Components, Packaging and Manufacturing Technology, 2019, 9(10): 1952-1962. DOI URL

芯片	尺寸/mm²	算力/TFLOPS	Fabric带宽/（TB·s^-1）	存储带宽/（TB·s^-1）	晶体管数/亿个
英伟达A100芯片	826	312	0.6	2.039	540
特斯拉Dojo训练Tile	>92903	9000	36	10	12500
提升倍数	112X	29X	60X	4.9X	23X

芯片	尺寸/mm²	算力/TFLOPS	Fabric带宽/（TB·s^-1）	存储带宽/（TB·s^-1）	晶体管数/亿个
英伟达A100芯片	826	312	0.6	2.039	540
特斯拉Dojo训练Tile	>92903	9000	36	10	12500
提升倍数	112X	29X	60X	4.9X	23X

Development and Opportunity of Advanced Packaging Technology

先进封装技术的发展与机遇

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 22

References 27

Related Articles 4

Recommended Articles

Metrics

Access Statistics

Links

Contact Us

WeChat

[1]	WANG Tongliang, YANG Menglu. Status and Future of Digital and Intelligent Development of Offshore Petroleum Engineering [J]. Science and Technology Foresight, 2023, 2(2): 105-120.
[2]	YAO Peng, SONG Changming, HU Yang, CAI Jian, YIN Shouyi, WU Huaqiang. Future Technical Development Approach for High Computing Power Chips [J]. Science and Technology Foresight, 2022, 1(3): 115-129.
[3]	LIU Weiping, WANG Zongyuan. Development Status and Trend of EDA and IP Core Industries [J]. Science and Technology Foresight, 2022, 1(3): 90-100.
[4]	CHEN Ge, YANG Jie, WU Lixin. Artificial Intelligence-aided Remote Sensing of the Intermediate Ocean [J]. Science and Technology Foresight, 2022, 1(2): 103-120.