CoChat
Sign Up
Cited by 94Photonic and Optical DevicesNeural Networks and Reservoir ComputingAdvanced Optical Network Technologies

The DOI matches its Crossref record; the references it cites are not checked yet.

Scaling up silicon photonic-based accelerators: Challenges and opportunities

Open access
Al-Qadasi, M. A., Chrostowski, L., Shastri, B. J., Shekhar, S.·2022·APL Photonics
DOI 10.1063/5.0070992
Open this paper in CoChat — free accountOpen PDF

Abstract

Digital accelerators in the latest generation of complementary metal–oxide–semiconductor processes support, multiply, and accumulate (MAC) operations at energy efficiencies spanning 10–100 fJ/Op. However, the operating speed for such MAC operations is often limited to a few hundreds of MHz. Optical or optoelectronic MAC operations on today’s SOI-based silicon photonic integrated circuit platforms can be realized at a speed of tens of GHz, leading to much lower latency and higher throughput. In this Perspective, we study the energy efficiency of integrated silicon photonic MAC circuits based on Mach–Zehnder modulators and microring resonators. We describe the bounds on energy efficiency and scaling limits for N × N optical networks with today’s technology based on the optical and electrical link budget. We also describe research directions that can overcome the current limitations.

Overview

Integrity signals

Registry

Crossref · 10.1063/5.0070992

Cites

Statistics

Recomputing p-values needs the paper’s full text.

Verification

Claims the paper makes

We couldn’t retrieve this paper’s full text, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Improving thermal insulation reduces metal heater power consumption by >10× at the expense of an even higher response time.

    “Improving thermal insulation reduces their power consumption by > 10 × at the expense of an even higher response time76.”

    From Roadmapping the next generation of silicon photonics (2024) · cites this paper as [76]

  • Such improvements are crucial to achieve accuracy comparable to the incumbent CMOS EIC compute engines.

    “Such improvements are crucial to achieve accuracy comparable to the incumbent CMOS EIC compute engines76.”

    From Roadmapping the next generation of silicon photonics (2024) · cites this paper as [76]

Citations

The citation lists haven’t been fetched yet

References 122 · Cited by 94

Research works printed but not linked: 86

  1. 1982, Why systolic architectures?, Computer, 15, 37, 10.1109/mc.1982.1653825 · DOI
  2. 2020, A dual-polarization silicon-photonic coherent transmitter supporting 552 Gb/s/wavelength, IEEE J. Solid-State Circuits, 55, 2597, 10.1109/jssc.2020.2988399 · DOI · Paper page
  3. 2016, Silicon-photonics microring links for datacenters—Challenges and opportunities, IEEE J. Sel. Top. Quantum Electron., 22, 194, 10.1109/jstqe.2016.2582345 · DOI
  4. 2017, A 40-Gb/s PAM-4 transmitter based on a ring-resonator optical DAC in 45-nm SOI CMOS, IEEE J. Solid-State Circuits, 52, 3503, 10.1109/jssc.2017.2748620 · DOI · Paper page
  5. 2019, Monitoring and automatic tuning and stabilization of a 2×2 MZI optical switch for large-scale WDM switch networks, Opt. Express, 27, 24747, 10.1364/oe.27.024747 · DOI
  6. 2019, Photoconductive heaters enable control of large-scale silicon photonic ring resonator circuits, Optica, 6, 84, 10.1364/optica.6.000084 · DOI
  7. 2020, Large-scale silicon photonics switch based on 45-nm CMOS technology, Opt. Commun., 466, 125677, 10.1016/j.optcom.2020.125677 · DOI
  8. 2019, Silicon photonic modulator neuron, Phys. Rev. Appl., 11, 064043, 10.1103/physrevapplied.11.064043 · DOI
  9. 2017, Deep learning with coherent nanophotonic circuits, Nat. Photonics, 11, 441, 10.1038/nphoton.2017.93 · DOI · Paper page
  10. 2018, Training of photonic neural networks through in situ backpropagation and gradient measurement, Optica, 5, 864, 10.1364/optica.5.000864 · DOI
  11. 2019, Digital electronics and analog photonics for convolutional neural networks (DEAP-CNNs), IEEE J. Sel. Top. Quantum Electron., 26, 7701213, 10.1109/JSTQE.2019.2945540 · DOI
  12. 2021, Silicon photonics: A brief tutorial, IEEE Solid-State Circuits Mag., 13, 22, 10.1109/mssc.2021.3088966 · DOI · Paper page
  13. 2022, Prospects and applications of photonic neural networks, Adv. Phys.: X, 7, 1981155, 10.1080/23746149.2021.1981155 · DOI
  14. M. J. Filipovich , Z.Guo, M.Al-Qadasi, B. A.Marquez, H. D.Morison, V. J.Sorger, P. R.Prucnal, S.Shekhar, and B. J.Shastri, “Monolithic silicon photonic architecture for training deep neural networks with direct feedback alignment,” arXiv:2111.06862 [cs.LG] (2021). · DOI
  15. 2018, Field-programmable photonic arrays, Opt. Express, 26, 27265, 10.1364/oe.26.027265 · DOI
  16. 2020, Photonic integrated field-programmable disk array signal processor, Nat. Commun., 11, 406, 10.1038/s41467-019-14249-0 · DOI
  17. 2020, Integrated coherent Ising machines based on self-phase modulation in microring resonators, IEEE J. Sel. Top. Quantum Electron., 26, 5900115, 10.1109/jstqe.2019.2929184 · DOI
  18. 2018, Large-scale silicon quantum photonics implementing arbitrary two-qubit processing, Nat. Photonics, 12, 534, 10.1038/s41566-018-0236-y · DOI · Paper page
  19. 1994, Experimental realization of any discrete unitary operator, Phys. Rev. Lett., 73, 58, 10.1103/physrevlett.73.58 · DOI · Paper page
  20. 2016, Optimal design for universal multiport interferometers, Optica, 3, 1460, 10.1364/optica.3.001460 · DOI · Paper page
  21. 2020, Parallel programming of an arbitrary feedforward photonic network, IEEE J. Sel. Top. Quantum Electron., 26, 6100813, 10.1109/jstqe.2020.2997849 · DOI
  22. 2019, 8 × 8 reconfigurable quantum photonic processor based on silicon nitride waveguides, Opt. Express, 27, 26842, 10.1364/oe.27.026842 · DOI
  23. 2015, Electro-optical co-simulation for integrated CMOS photonic circuits with VerilogA, Opt. Express, 23, 27180, 10.1364/oe.23.027180 · DOI
  24. 2018, In situ 3D nanoprinting of free-form coupling elements for hybrid photonic integration, Nat. Photonics, 12, 241, 10.1038/s41566-018-0133-4 · DOI
  25. 1986, 1.0 dB low-loss coupling of laser diode to single-mode fibre using a planoconvex graded-index rod lens, Electron. Lett., 22, 1122, 10.1049/el:19860769 · DOI
  26. 2018, Wire-bonding assembly, Nat. Photonics, 12, 500, 10.1038/s41566-018-0251-z · DOI
  27. 2012, Photonic wire bonding: A novel concept for chip-scale interconnects, Opt. Express, 20, 17667, 10.1364/oe.20.017667 · DOI
  28. 2015, Connecting silicon photonic circuits to multicore fibers by photonic wire bonding, J. Lightwave Technol., 33, 755, 10.1109/jlt.2014.2373051 · DOI · Paper page
  29. 2019, Silicon photonic circuit design using rapid prototyping foundry process design kits, IEEE J. Sel. Top. Quantum Electron., 25, 8201326, 10.1109/jstqe.2019.2917501 · DOI
  30. 1997, Silicon optical bench waveguide technology, 319
  31. 2016, A silicon photonic PAM-4 modulator based on dual-parallel Mach–Zehnder interferometers, IEEE Photonics J., 8, 7800610, 10.1109/jphot.2015.2512105 · DOI
  32. 2017, Mode/polarization manipulation in silicon photonics, J. Phys.: Conf. Ser., 844, 012039, 10.1088/1742-6596/844/1/012039 · DOI
  33. 2019, Polarization- and wavelength-agnostic nanophotonic beam splitter, Sci. Rep., 9, 3604, 10.1038/s41598-019-40497-7 · DOI
  34. 2015, Broadband silicon photonic directional coupler using asymmetric-waveguide based phase control, Opt. Express, 23, 3795, 10.1364/oe.23.003795 · DOI
  35. 2016, Low-loss and broadband 2 × 2 silicon thermo-optic Mach–Zehnder switch with bent directional couplers, Opt. Lett., 41, 836, 10.1364/ol.41.000836 · DOI
  36. 2020, Introduction to Fiber-Optic Communications
  37. 2016, High-speed frequency modulated low-noise single-frequency fiber laser, IEEE Photonics Technol. Lett., 28, 1692, 10.1109/lpt.2016.2529652 · DOI
  38. 2019, Optimization of thermo-optic phase-shifter design and mitigation of thermal crosstalk on the SOI platform, Opt. Express, 27, 10456, 10.1364/oe.27.010456 · DOI
  39. 2014, Efficient, compact and low loss thermo-optic phase shifter in silicon, Opt. Express, 22, 10487, 10.1364/oe.22.010487 · DOI · Paper page
  40. 2012, 10-Gbps, 5.3-mW optical transmitter and receiver circuits in 40-nm CMOS, IEEE J. Solid-State Circuits, 47, 2049, 10.1109/jssc.2012.2197234 · DOI
  41. 2009, A single-chip CMOS-based parallel optical transceiver capable of 240-Gb/s bidirectional data rates, J. Lightwave Technol., 27, 915, 10.1109/jlt.2008.927759 · DOI
  42. 2019, A process and temperature insensitive CMOS linear TIA for 100 Gb/s/λ PAM-4 optical links, IEEE J. Solid-State Circuits, 54, 3180, 10.1109/jssc.2019.2939652 · DOI
  43. 2017, A 56-Gb/s PAM4 wireline transceiver using a 32-way time-interleaved SAR ADC in 16-nm FinFET, IEEE J. Solid-State Circuits, 52, 1101, 10.1109/jssc.2016.2632300 · DOI
  44. 2002, OC-192 transmitter and receiver in standard 0.18-/spl mu/m CMOS, IEEE J. Solid-State Circuits, 37, 1768, 10.1109/jssc.2002.804336 · DOI
  45. 2018, Ultralow-power (1.59 mW/Gbps), 56-Gbps PAM4 operation of Si photonic transmitter integrating segmented PIN Mach–Zehnder modulator and 28-nm CMOS driver, J. Lightwave Technol., 36, 1275, 10.1109/jlt.2018.2799965 · DOI
  46. 2020, A sub-pJ/bit, low-ER Mach–Zehnder-based transmitter for chip-to-chip optical interconnects, IEEE J. Sel. Top. Quantum Electron., 26, 8301910, 10.1109/jstqe.2019.2954705 · DOI
  47. 2020, Photonic multiply-accumulate operations for neural networks, IEEE J. Sel. Top. Quantum Electron., 26, 7701518, 10.1109/jstqe.2019.2941485 · DOI
  48. 2016, Energy scaling advantages of resistive memory crossbar based computation and its application to sparse coding, Front. Neurosci., 9, 484, 10.3389/fnins.2015.00484 · DOI
  49. 2014, Broadcast and weight: An integrated network for scalable photonic spike processing, J. Lightwave Technol., 32, 4029, 10.1109/jlt.2014.2345652 · DOI
  50. 2021, A 3-D-integrated silicon photonic microring-based 112-Gb/s PAM-4 transmitter with nonlinear equalization and thermal control, IEEE J. Solid-State Circuits, 56, 19, 10.1109/jssc.2020.3022851 · DOI
  51. 2019, Ultra-compact silicon photonics switch with high-density thermo-optic heaters, Opt. Express, 27, 10332, 10.1364/oe.27.010332 · DOI
  52. 2014, A high-speed, tunable silicon photonic ring modulator integrated with ultra-efficient active wavelength control, Opt. Express, 22, 12628, 10.1364/oe.22.012628 · DOI
  53. 2016, Microring weight banks, IEEE J. Sel. Top. Quantum Electron., 22, 312, 10.1109/jstqe.2016.2573583 · DOI
  54. 2016, Crosstalk in SOI microring resonator-based filters, J. Lightwave Technol., 34, 2886, 10.1109/jlt.2015.2480101 · DOI · Paper page
  55. 2008, Silicon microring resonators with 1.5-μm radius, Opt. Express, 16, 4309, 10.1364/oe.16.004309 · DOI
  56. 2019, Compact wavelength- and bandwidth-tunable microring modulator, Opt. Express, 27, 26661, 10.1364/oe.27.026661 · DOI
  57. 2016, FSR-free silicon-on-insulator microring resonator based filter with bent contra-directional couplers, Opt. Express, 24, 29009, 10.1364/oe.24.029009 · DOI
  58. 2016, A novel approach to photonic packaging leveraging existing high-throughput microelectronic facilities, IEEE J. Sel. Top. Quantum Electron., 22, 455, 10.1109/jstqe.2016.2593637 · DOI
  59. 2020, Deep neural network through an InP SOA-based photonic integrated cross-connect, IEEE J. Sel. Top. Quantum Electron., 26, 7701111, 10.1109/jstqe.2019.2945548 · DOI
  60. 2016, 25 Gbps low-voltage waveguide Si–Ge avalanche photodiode, Optica, 3, 793, 10.1364/optica.3.000793 · DOI
  61. 2010, High-sensitivity 10 Gbps Ge-on-Si photoreceiver operating at λ 1.55 μm, Opt. Express, 18, 16474, 10.1364/oe.18.016474 · DOI
  62. 2012, Power-efficient III-V/silicon external cavity DBR lasers, Opt. Express, 20, 23456, 10.1364/oe.20.023456 · DOI
  63. 2012, High-output-power, single-wavelength silicon hybrid laser using precise flip-chip bonding technology, Opt. Express, 20, 28057, 10.1364/oe.20.028057 · DOI
  64. 2011, Optical isolator using two tandem phase modulators, Opt. Lett., 36, 4293, 10.1364/ol.36.004293 · DOI
  65. 2010, Thermally tunable silicon racetrack resonators with ultralow tuning power, Opt. Express, 18, 20298, 10.1364/oe.18.020298 · DOI
  66. 2016, A 45 nm CMOS-SOI monolithic photonics platform with bit-statistics-based resonant microring thermal tuning, IEEE J. Solid-State Circuits, 51, 893, 10.1109/jssc.2016.2519390 · DOI
  67. 2015, Robust postfabrication trimming of ultracompact resonators on silicon on insulator with relaxed requirements on resolution and alignment, Opt. Lett., 40, 4476, 10.1364/ol.40.004476 · DOI
  68. 2018, Germanium implanted photonic devices for post-fabrication trimming and programmable circuits, Proc. SPIE, 10823, 108230U, 10.1117/12.2503224 · DOI
  69. 2020, Performance analysis of a silicon NOEMS device applied as an optical modulator based on a slot waveguide, Opt. Express, 28, 38206, 10.1364/oe.411933 · DOI
  70. 2015, Digitally controlled phase shifter using an SOI slot waveguide with liquid crystal infiltration, IEEE Photonics Technol. Lett., 27, 1269, 10.1109/lpt.2015.2416438 · DOI
  71. 2018, Broadband nonvolatile photonic switching based on optical phase change materials: Beyond the classical figure-of-merit, Opt. Lett., 43, 94, 10.1364/ol.43.000094 · DOI
  72. 2019, Design of a compact low-loss phase shifter based on optical phase change material, IEEE Photonics Technol. Lett., 31, 1757, 10.1109/lpt.2019.2946187 · DOI
  73. 2015, Integrated all-photonic non-volatile multi-level memory, Nat. Photonics, 9, 725, 10.1038/nphoton.2015.182 · DOI
  74. 2011, Semiconductor Optical Amplifiers
  75. 2018, All optical switching and associated technologies: A review, J. Opt., 47, 307, 10.1007/s12596-018-0452-3 · DOI
  76. 2000, Theory and measurement techniques for the noise figure of optical amplifiers, Opt. Fiber Technol., 6, 122, 10.1006/ofte.2000.0327 · DOI
  77. 2010, Semiconductor optical amplifiers, 233
  78. 2018, ‘Cross gain modulation’ effect of SOAs for different modulation formats, Int. J. Eng. Res. Technol., 2, 357
  79. 2004, A dispersion management scheme for reducing SOA-induced crosstalk in WDM links, J. Lightwave Technol., 22, 417, 10.1109/jlt.2003.822117 · DOI
  80. 2021, Photonics for artificial intelligence and neuromorphic computing, Nat. Photonics, 15, 102, 10.1038/s41566-020-00754-y · DOI · Paper page
  81. 2009, High-speed and low-power electro-optical DSP coprocessor, J. Opt. Soc. Am. A, 26, A11, 10.1364/josaa.26.000a11 · DOI
  82. W. Zhang , C.Huang, S.Bilodeau, A.Jha, E.Blow, T. F. D.Lima, B. J.Shastri, and P.Prucnal, “Microring weight banks control beyond 8.5-bits accuracy,” arXiv:2104.01164 [physics.app-ph] (2021). · arXiv
  83. I. Hubara , M.Courbariaux, D.Soudry, R.El-Yaniv, and Y.Bengio, “Quantized neural networks: Training neural networks with low precision weights and activations,” arXiv:1609.07061 [cs.NE] (2016). · arXiv
  84. 2016, Convolutional networks for fast, energy-efficient neuromorphic computing, Proc. Natl. Acad. Sci. U. S. A., 113, 11441, 10.1073/pnas.1604850113 · DOI
  85. 2021, Optical and electrical programmable computing energy use comparison, Opt. Express, 29, 13153, 10.1364/oe.420027 · DOI
  86. 2016, Photonic packaging: Transforming silicon photonic integrated circuits into photonic devices, Appl. Sci., 6, 426, 10.3390/app6120426 · DOI

Metadata

Authors
Al-Qadasi, M. A., Chrostowski, L., Shastri, B. J., & Shekhar, S.
Affiliations
  • University of British Columbia (CA)
  • Queen's University (CA)
Corresponding author
Al-Qadasi, M. A.
ORCID
Venue
APL Photonics
Publisher
AIP Publishing
Published
2022-02-01
Volume
7
Issue
2
ISSN
2378-0967 (ISSN-L)
DOI
10.1063/5.0070992
OpenAlex
W4210884760
Cited by
94
References
122
Citations by year
2026: 17 · 2025: 17 · 2024: 23 · 2023: 24 · 2022: 13
FWCI
5.27
Type
journal article
Language
en
Open access
Licence
Locations
  • APL Photonics (publishedVersion, cc-by)
  • arXiv (Cornell University) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
  • arXiv (Cornell University)
Funders
Natural Sciences and Engineering Research Council of Canada
Topics
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Photonic and Optical Devices
  • Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Reservoir Computing
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Optical Network Technologies
Keywords
silicon photonics, multiply-accumulate operations, energy efficiency, Mach-Zehnder modulator, microring resonators, scaling limits, optical networks, link budget, SOI photonics, throughput and latency

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review