Large-Scale Cryogenic Integration Approach for Superconducting High-Performance Computing

被引:26
|
作者
Das, Rabindra N. [1 ]
Bolkhovsky, Vladimir [1 ]
Tolpygo, Sergey K. [1 ]
Gouker, Pascale [1 ]
Johnson, Leonard M. [1 ]
Dauler, Eric A. [1 ]
Gouker, Mark A. [1 ]
机构
[1] MIT, Lincoln Lab, Quantum Informat & Integrated Nanosyst Grp, Lexington, MA 02420 USA
关键词
S-MCM; microbump; packaging architecture; flip-chip; ELECTRONICS;
D O I
10.1109/ECTC.2017.54
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Superconducting single-flux-quantum-based (SFQ) digital integrated circuits (ICs) are a promising candidate for highspeed and ultralow energy dissipation computing systems. Circuits based on several versions of SFQ-type logic, RQL, and AQFP logic have been demonstrated with complexities reaching up to a few tens of thousands of gates. Packaging many superconducting ICs using microbump (15 mu m or less) technology and performing high throughput, nearly lossless data transfer between various superconducting and/or CMOS chips are highly desirable for a hybrid superconducting computer architecture, but this density has not yet been demonstrated. An efficient way to achieve this goal is to couple the chips through a passive superconductive multichip module (S-MCM) that distributes information between integrated circuits utilizing lossless superconducting transmission lines. Here we show implementation of such a superconducting base using well-defined impedance lines to couple multiple ICs to enable a cryogenic integration approach for possible future hybrid superconducting computing systems. The use of indium-based microbumps to form interconnects between an S-MCM and superconducting ICs, and its electrical performance, are discussed. Optimized microbumps were used to fabricate interconnections on a large S-MCM by using thermocompression(TC) bonding. From 17200 to 68800 microbumps fabricated in a 5 mm x 5 mm active area of an S-MCM, having pitches ranging from 35 mu m to 15 mu m, were bonded with a superconducting IC to create a daisy chain structure. In addition, a large active area (10 mm x 10 mm) flipchip having continuous niobium daisy chains with up to 77,500 bumps with 15 mu m bump diameter and 35 mu m pitch was also demonstrated. As a case study, an optimized microbump interconnect construction for attaching 16 superconducting chips with a large 32 mm x 32 mm superconducting base was fabricated and tested at room temperature and 4.2 K to study the structural and electrical integrity. Josephson junctions and niobium lines on integrated superconducting chips and the superconducting base maintained their I-V characteristics, which allows the design of building blocks for a superconducting computing system.
引用
收藏
页码:675 / 683
页数:9
相关论文
共 50 条
  • [41] A novel approach for hybrid performance modelling and prediction of large-scale computing systems
    Pllana, Sabri
    Benkner, Siegfried
    Xhafa, Fatos
    Barolli, Leonard
    INTERNATIONAL JOURNAL OF GRID AND UTILITY COMPUTING, 2009, 1 (04) : 316 - 327
  • [42] AI Meets Exascale Computing: Advancing Cancer Research With Large-Scale High Performance Computing
    Bhattacharya, Tanmoy
    Brettin, Thomas
    Doroshow, James H.
    Evrard, Yvonne A.
    Greenspan, Emily J.
    Gryshuk, Amy L.
    Hoang, Thuc T.
    Lauzon, Carolyn B. Vea
    Nissley, Dwight
    Penberthy, Lynne
    Stahlberg, Eric
    Stevens, Rick
    Streitz, Fred
    Tourassi, Georgia
    Xia, Fangfang
    Zaki, George
    FRONTIERS IN ONCOLOGY, 2019, 9
  • [43] High-performance Placement for Large-scale Heterogeneous FPGAs with Clock Constraints
    Zhu, Ziran
    Mei, Yangjie
    Li, Zijun
    Lin, Jingwen
    Chen, Jianli
    Yang, Jun
    Chang, Yao-Wen
    PROCEEDINGS OF THE 59TH ACM/IEEE DESIGN AUTOMATION CONFERENCE, DAC 2022, 2022, : 643 - 648
  • [44] IGD: high-performance search for large-scale genomic interval datasets
    Feng, Jianglin
    Sheffield, Nathan C.
    BIOINFORMATICS, 2021, 37 (01) : 118 - 120
  • [45] High-performance large-scale simulation of multi-stable metastructures
    Hwang, Myungwon
    Scalo, Carlo
    Arrieta, Andres F.
    COMPUTER PHYSICS COMMUNICATIONS, 2022, 277
  • [46] A High-Performance, Scalable Infrastructure for Large-Scale Active DNS Measurements
    van Rijswijk-Deij, Roland
    Jonker, Mattijs
    Sperotto, Anna
    Pras, Aiko
    IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, 2016, 34 (06) : 1877 - 1888
  • [47] High-Performance Large-Scale Atomistic Simulation of Thin Films Deposition
    F. V. Grigoriev
    V. B. Sulimov
    A. V. Tikhonravov
    Lobachevskii Journal of Mathematics, 2022, 43 : 857 - 863
  • [49] Data Centric Framework for Large-scale High-performance Parallel Computation
    Ono, Kenji
    Kawashima, Yasuhiro
    Kawanabe, Tonaohiro
    2014 INTERNATIONAL CONFERENCE ON COMPUTATIONAL SCIENCE, 2014, 29 : 2336 - 2350
  • [50] SciAP: A Programmable, High-Performance Platform for Large-Scale Scientific Data
    Tian, Yang
    Li, Chao
    Liu, Chao
    Yan, Haihua
    2018 INTERNATIONAL CONFERENCE ON CLOUD COMPUTING, BIG DATA AND BLOCKCHAIN (ICCBB 2018), 2018, : 148 - 154