Pith. sign in

REVIEW 4 major objections 5 minor 52 references

A Mixed-Signal Photonic SRAM-based High-Speed Energy-Efficient Photonic Tensor Core with Novel Electro-Optic ADC

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read The paper claims a photonic tensor core whose weights live in differential photonic SRAM bitcells and whose outputs are digitized by a one-hot electro-optic ADC can do matrix multiply at 4.10 TOPS with 3.02 TOPS/W efficiency on a 45 nm…

desk verdict A coherent photonic tensor-core design with a genuinely new electro-optic ADC, but the headline TOPS/W rests on unexamined thermal margins and a nonstandard TOPS definition. read the letter →

arxiv 2506.22705 v1 pith:R5WGGRR4 submitted 2025-06-28 physics.optics cs.SYeess.SY

classification physics.opticscs.SYeess.SY
keywords photonicin-memorycomputingSRAMmicroringresonatorelectro-opticADCtensorcorewavelengthdivisionmultiplexingmixed-signalmatrixmultiplication
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper designs and simulates a photonic tensor core that stores matrix weights in a differential photonic SRAM bitcell built from two cross-coupled microring resonators and four photodiodes. A wavelength-multiplexed array of microrings multiplies intensity-encoded analog inputs by those stored binary weights and sums the products optically at photodetectors. A one-hot electro-optic ADC converts the summed photocurrents into digital codes, claiming 8 GS/s sampling at 2.32 pJ per conversion. With all stages using fabrication-friendly silicon photonics in a monolithic 45 nm process, the core reports 4.10 TOPS throughput, 3.02 TOPS/W power efficiency, and 20 GHz weight-update speed. A sympathetic reader would take the contribution to be an end-to-end mixed-signal architecture that removes the usual off-chip electrical digitization bottleneck.

What carries the argument

The load-bearing mechanism is the cross-coupled differential photonic SRAM bitcell: two microrings (M1 and M2) with thru and drop ports connected to photodiodes form a latch whose storage nodes Q and QB tune one ring into resonance with the input laser and the other out of resonance, so a stored weight is also a physical multiplier of an intensity-encoded input. The second load-bearing mechanism is the 1-hot electro-optic ADC: for a $p$-bit conversion, $2^p$ microrings are biased with reference voltages that tile the input range, and the voltage-dependent resonance notch means only the ring assigned to the current input window transmits below the reference power; a balanced photodiode pair, a transimpedance amplifier, and a ROM-based ceiling decoder turn that single activation into a digital code. Together these mechanisms let weights, multiplication, and digitization all live in the optical domain while remaining compatible with standard silicon photonic fabrication.

What would settle it

Run a temperature sweep or a foundry-process Monte Carlo simulation on the 3-bit eoADC and measure each ring's thru-port crossing voltage: if any ring's dip boundary shifts by more than roughly a quarter of one LSB code width, more than one thresholding block activates (or none does) and the reported transfer function and DNL in the paper no longer hold at 8 GS/s.

Watch

Extended reading notes

Core claim

The central claim is that an end-to-end photonic tensor core can be assembled from the same microring resonators and photodiodes used for memory, multiplication, and analog-to-digital conversion. The pSRAM bitcell latches a weight by holding complementary voltages that tune one ring into resonance and the other out of resonance, so each bitcell directly gates an incoming laser intensity. The WDM compute core multiplies a $1\times 4$ analog input vector by 3-bit weights per row, using four wavelengths spaced 2.33 nm apart within a 9.36 nm free spectral range, and combines results through photodiode current summation. The eoADC tiles the input voltage range across eight rings; only the ring assigned to the current voltage window drops its thru-port power below an optical reference, activating a single thresholding block, and a ROM-based ceiling decoder emits the corresponding binary code even when the input sits on a code boundary. The paper reports end-to-end numbers of 4.10 TOPS, 3.02 TOPS/W, 8 GS/s at 2.32 pJ per conversion, and a 20 GHz weight-update rate, with each wavelength channel simulated separately and combined linearly because the process design kit simulates one wavelength at a time.

Load-bearing premise

The one-hot ADC works only if each microring's resonance dip stays aligned to its assigned input-voltage window to well under one code width; the paper relies on thermal tuning to hold this alignment and includes no temperature-sweep, process-variation, or noise analysis that shows it holds.

Editorial extensions

If this is right

  • Weight updates at 20 GHz would make frequent in-situ retraining or streaming weight refresh practical, a capability the comparison table contrasts with sub-0.5 GHz FPGA-controlled and slow PCM-based weight banks.
  • Because the eoADC does one-hot conversion at flash-ADC speeds, the end-to-end core avoids off-chip power measurement or electrical ADC bottlenecks that limit earlier photonic in-memory macros.
  • The WDM compute macro can be replicated to extend $1\times 4$ vector multiplies to $1\times 16$ and $m\times n$ matrices by summing photodiode currents, so the architecture is scalable without changing the bitcell.
  • The ceiling-priority ROM decoder prevents two codes from firing at a boundary input, so conversion at the midpoints between adjacent windows remains deterministic.
  • Higher precision than 3 bits can be reached by optimizing rings or cascading lower-bit ADCs with shift-and-add, as the paper states for the eoADC.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A direct experimental check would be simultaneous four-wavelength operation with all channels powered at once, since the current verification simulates each wavelength separately and linearly superposes photocurrents; nonlinear optical effects or thermal crosstalk between channels would show up only in that measurement.
  • The one-hot ADC principle is a parallel bank of optical comparators, so time-interleaving several eoADC slices could push sampling rates beyond the reported 8 GS/s, a route the paper mentions without quantifying the added power or timing-skew cost.
  • Because the eoADC's accuracy is set by how sharply each ring's notch moves with voltage, improving MRR Q-factor or voltage-modulation efficiency would trade speed for bit precision; measuring that trade-off per code width would tell whether 4-bit or 5-bit conversions are practical in the same node.
  • The reported 3.02 TOPS/W depends on the 0.23 laser wall-plug efficiency; a system with a different laser or on-chip laser integration would need that number re-derived from the optical power budget.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The manuscript proposes a mixed-signal photonic tensor core built around a differential photonic SRAM (pSRAM) bitcell for weight storage, microring-resonator (MRR)-based wavelength-division-multiplexed vector multiplication, and a new one-hot encoding electro-optic ADC (eoADC) that uses MRR transmission dips as voltage comparators. The authors claim a 20 GHz weight-update rate, an 8 GS/s eoADC with 2.32 pJ per conversion, and, for a 16x16 core at 3-bit weight precision, a throughput of 4.10 TOPS and an efficiency of 3.02 TOPS/W, all in the GlobalFoundries 45SPCLO monolithic node. The evidence is component-level: pSRAM write transients, MRR transmission spectra, vector-multiplication transients, ADC transfer and DNL curves, and a brief performance comparison table.

Significance. If the claims are substantiated, the pSRAM bitcell and the one-hot eoADC concept are interesting contributions: the former provides an optically writable, electrically readable weight cell with a fast update path, and the latter is a genuinely different ADC approach that could reduce comparator power. The use of a specific foundry process and the inclusion of laser wall-plug efficiency in some component estimates are positive features. However, the central performance claims are not yet supported at the system level: the eoADC's one-hot operation depends on resonance alignment that is not analyzed under temperature or process variation, the TOPS count is not derived, and heater power is omitted from the energy numbers. These are correctable with additional analysis, and the paper would be suitable for a major revision if that analysis is supplied.

major comments (4)
  1. [Section II-C, Fig. 10, and Section IV-C] The one-hot eoADC requires each MRR's transmission dip to remain centered in its assigned 225 mV code window. The paper provides no temperature sweep, process-corner/Monte Carlo analysis, or noise analysis for the MRRs or the thresholding blocks, and the Introduction's statement that thermal fluctuations can be mitigated with integrated heaters does not address process variation or quantify heater power. A resonance shift of more than about half an LSB will produce either a missing code (no block activates) or double activation for a non-boundary input; the ROM ceiling decoder cannot recover from arbitrary misalignment. Because reliable 8 GS/s conversion is the advertised speed limiter of the tensor core, this omission is load-bearing for both the throughput and the efficiency claims.
  2. [Section IV-D] The headline values of 4.10 TOPS and 3.02 TOPS/W are stated without a derivation. The sentence '1 operation = 3-bit multiplication/addition' does not specify how many operations occur per conversion, what clock rate is assumed, or how the 16x16 core and 768 pSRAM cells map onto that count. As a sanity check, a 16x16 matrix-vector product performs 256 MACs per cycle; at 8 GS/s that would be about 2 TOPS under a MAC-based convention, so the claimed 4.10 TOPS implies a different operation-counting rule that must be stated explicitly. The authors should provide a step-by-step calculation of throughput and total power, including all components listed in Section IV-D.
  3. [Sections IV-B and IV-D] The verification is component-level and not end-to-end. The paper explicitly states that each WDM wavelength channel is simulated separately and results are combined linearly, and the ADC, vector-multiplication core, pSRAM, and TIA are not co-simulated. A claim that the assembled system achieves 4.10 TOPS at 3.02 TOPS/W therefore rests on an unvalidated linear combination of separately simulated blocks. The authors should either provide an end-to-end system simulation or clearly label the headline numbers as projections with a list of all extrapolation steps.
  4. [Section IV-C and Section IV-D] Thermal tuning power is excluded from the energy accounting. The paper acknowledges in the Introduction that MRRs are susceptible to thermal fluctuations and cites integrated heaters as the mitigation, but the 2.32 pJ/conversion and the 3.02 TOPS/W figures do not include the electrical power needed to hold the compute MRRs and the eight eoADC MRRs on their assigned resonance wavelengths. For a silicon-photonics process, temperature-induced resonance drift is large enough that this power is likely non-negligible. The authors should estimate the per-ring tuning power and include it in the efficiency calculation, or show that the eoADC and compute rings are stable without heating.
minor comments (5)
  1. [Section IV-A] The claim that the pSRAM consumes 0.5 pJ per switching event at 20 GHz follows from the -20 dBm optical bias and 0.23 wall-plug efficiency, but the text should state explicitly whether write-pulse energy, photodiode bias, and driver energy are included in that number.
  2. [Section IV-C] The phrase 'total optical power is 7.58 mW' is ambiguous when combined with 'wall-plug efficiency of 0.23': it should be stated whether 7.58 mW is the optical power at the laser output or the electrical input power to the laser, because the 2.32 pJ/conversion figure depends on which convention is used.
  3. [Section II-C and Fig. 9] For the 2 V boundary input, two thresholding blocks (B4 and B5) activate and the decoder outputs 100; the text should explain how the 'ceiling' ROM resolves two active blocks and how this maps to the ideal code, since this is the claimed robustness mechanism.
  4. [Section III and Section IV-B] The FSR and channel spacing are given as 9 nm and 2 nm in Section III but as 9.36 nm and 2.33 nm in Section IV-B; these numbers should be made consistent.
  5. [Table I] The 'Weight Update' column mixes units (GHz, Hz, and qualitative entries); a fair comparison would state the update rate in operations per second or Hz for every row and would specify what is being updated.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: headline TOPS/TOPS/W and eoADC energy figures are assembled from component simulations and external efficiency references, not from the target result itself.

full rationale

The paper's performance claims are computed from independent component-level simulations rather than from the target result. Section IV-C derives the eoADC's 2.32 pJ/conversion by dividing the simulated total optical and electrical power (7.58 mW and 11 mW) by the 8 GS/s sampling rate. Section IV-D derives 4.10 TOPS from the 16x16 core's 256 multiply-add operations times two ops per MAC times the 8 GS/s ADC-limited rate, and 3.02 TOPS/W by including pSRAM bitcells, compute cores, eoADC, TIA [52], and laser wall-plug efficiency [47]. The eoADC reference voltages are design parameters chosen to place MRR resonances at the desired code boundaries; the simulated transfer function and DNL verify that the design behaves as specified, rather than fitting a parameter to a target prediction. Self-citations [44] and [45] provide background for the pSRAM bitcell and binary-scaled power splitters, but those components are also described and simulated in the paper itself, so the self-citations are not load-bearing. Concerns about thermal drift, process variation, and tuning power are robustness or correctness risks, not circular reasoning.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

The central metrics depend on several hand-set design values (optical powers, supply, channel count, MRR geometry, reference voltages) and on external references for laser efficiency and TIA performance. None of these are fitted to experimental data in this paper, so the ledger mainly exposes assumptions rather than circularity. The TOPS counting convention is the most consequential free choice because it changes the headline throughput depending on how operations are counted.

free parameters (8)
  • pSRAM write pulse width and bias laser power = 50 ps, -20 dBm
    Chosen for simulation; together they set the claimed 20 GHz update rate and 0.5 pJ switching energy.
  • eoADC input and reference optical power = 200 uW input, 18 uW reference per channel
    Hand-selected to produce one-hot dips; total optical power and 2.32 pJ per conversion depend on them.
  • eoADC supply voltage = 1.8 V
    Sets the electrical power of 11 mW and affects the conversion energy calculation.
  • Laser wall-plug efficiency = 0.23
    Taken from external VCSEL reference [47]; used to convert optical power into electrical power for the TOPS/W estimate.
  • Wavelength channels per vector macro = 4 channels at 2.33 nm spacing
    Design choice based on the 9.36 nm FSR; determines vector size and throughput.
  • MRR geometry = 7.5 um radius/200 nm gap for compute, 10 um radius/250 nm gap for ADC
    Chosen to achieve four resonance wavelengths and one-hot voltage windows; the simulated responses depend on these values.
  • eoADC reference voltage set = eight VREF values covering 0 to VFS
    Hand-set to segment the transfer function into eight 3-bit codes; determines DNL and code width.
  • TOPS counting convention = 1 operation = 3-bit multiplication/addition
    Nonstandard counting; the 4.10 TOPS figure is computed under this definition and is not directly comparable to conventional MAC TOPS.
assumptions (5)
  • domain assumption The cross-coupled pSRAM bitcell holds its state as long as optical bias and VDD are applied.
    Section II-A describes positive feedback; Fig. 5 verifies one write and hold sequence but includes no stability margin analysis.
  • domain assumption Photodiode currents from different wavelengths and compute macros sum linearly with negligible crosstalk at 2.33 nm channel spacing.
    Section III relies on current summation; crosstalk is asserted from FSR and spacing, not measured across all channels.
  • domain assumption MRR resonances can be thermally stabilized with integrated heaters.
    Mentioned in the Introduction with refs [37], [38]; no thermal simulation or tuning power is included in the energy estimates.
  • domain assumption The ROM-based ceiling decoder resolves boundary inputs without static current or metastability.
    Section II-C describes the ceiling function; only the 2 V boundary case is shown in Fig. 9.
  • domain assumption The cited TIA and cascaded amplifiers operate at 8 GS/s when connected to these photodiodes and MRRs.
    The TIA is cited from an external design [52]; the full ADC chain is not simulated together with a realistic photodiode model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Mixed-Signal Photonic SRAM-based High-Speed Energy-Efficient Photonic Tensor Core with Novel Electro-Optic ADC." pith.science (2026). https://pith.science/paper/R5WGGRR4

@misc{pith2026250622705,
  author       = {Pith},
  title        = {Pith review of: A Mixed-Signal Photonic SRAM-based High-Speed Energy-Efficient Photonic Tensor Core with Novel Electro-Optic ADC},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R5WGGRR4}},
  note         = {Machine review of arXiv:2506.22705}
}
read the original abstract

The rapid surge in data generated by Internet of Things (IoT), artificial intelligence (AI), and machine learning (ML) applications demands ultra-fast, scalable, and energy-efficient hardware, as traditional von Neumann architectures face significant latency and power challenges due to data transfer bottlenecks between memory and processing units. Furthermore, conventional electrical memory technologies are increasingly constrained by rising bitline and wordline capacitance, as well as the resistance of compact and long interconnects, as technology scales. In contrast, photonics-based in-memory computing systems offer substantial speed and energy improvements over traditional transistor-based systems, owing to their ultra-fast operating frequencies, low crosstalk, and high data bandwidth. Hence, we present a novel differential photonic SRAM (pSRAM) bitcell-augmented scalable mixed-signal multi-bit photonic tensor core, enabling high-speed, energy-efficient matrix multiplication operations using fabrication-friendly integrated photonic components. Additionally, we propose a novel 1-hot encoding electro-optic analog-to-digital converter (eoADC) architecture to convert the multiplication outputs into digital bitstreams, supporting processing in the electrical domain. Our designed photonic tensor core, utilizing GlobalFoundries' monolithic 45SPCLO technology node, achieves computation speeds of 4.10 tera-operations per second (TOPS) and a power efficiency of 3.02 TOPS/W.

Figures

Figures reproduced from arXiv: 2506.22705 by the authors.

Figure 1
Figure 1. Differential cross-coupled photonic SRAM bitcell. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Mixed-Signal multi-bit photonic vector multiplication compute core. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. (a) MRR transmission spectra as a function of the pn junction voltage, [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Mixed-signal multi-bit scalable 2D photonic tensor core enabling [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Verification of weight configuration in pSRAM bitcell. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Transmission spectra of the MRR as a function the ring adjustment [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 9
Figure 9. Figure 9: Transient verification results of the eoADC architecture. [PITH_FULL_IMAGE:figures/full_fig_p006_9.png]
Figure 10
Figure 10. Figure 10: ADC transfer function (left-subplot) and differential nonlinearity [PITH_FULL_IMAGE:figures/full_fig_p006_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 52 canonical work pages

  1. [22]

    A 7-nm compute-in-memory sram macro sup- porting multi-bit input, weight and output and achieving 351 tops/w and 372.4 gops,

    M. E. Sinangil et al. , “A 7-nm compute-in-memory sram macro sup- porting multi-bit input, weight and output and achieving 351 tops/w and 372.4 gops,” IEEE Journal of Solid-State Circuits , vol. 56, no. 1, pp. 188–198, 2020

  2. [1]

    Hitting the memory wall: Implications of the obvious,

    W. A. Wulf et al. , “Hitting the memory wall: Implications of the obvious,” ACM SIGARCH computer architecture news , vol. 23, no. 1, pp. 20–24, 1995

  3. [2]

    Memory devices and applications for in-memory computing,

    A. Sebastian et al. , “Memory devices and applications for in-memory computing,” Nature nanotechnology, vol. 15, no. 7, pp. 529–544, 2020

  4. [3]

    In-memory computing: Advances and prospects,

    N. Verma et al., “In-memory computing: Advances and prospects,”IEEE Solid-State Circuits Magazine , vol. 11, no. 3, pp. 43–55, 2019

  5. [4]

    A review on memristive devices and appli- cations,

    T. Prodromakis et al. , “A review on memristive devices and appli- cations,” in 2010 17th IEEE international conference on electronics, circuits and systems . IEEE, 2010, pp. 934–937

  6. [5]

    Integrated photonic computing beyond the von neu- mann architecture,

    X.-Y . Xu et al. , “Integrated photonic computing beyond the von neu- mann architecture,” ACS Photonics, vol. 10, no. 4, pp. 1027–1036, 2023

  7. [6]

    A survey on quantum computing technology,

    L. Gyongyosi et al. , “A survey on quantum computing technology,” Computer Science Review , vol. 31, pp. 51–71, 2019

  8. [7]

    Challenges and trends of sram-based computing-in- memory for ai edge devices,

    C.-J. Jhang et al., “Challenges and trends of sram-based computing-in- memory for ai edge devices,”IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 68, no. 5, pp. 1773–1786, 2021

Show all 52 references
  1. [8]

    Xnor-sram: In-memory computing sram macro for binary/ternary deep neural networks,

    S. Yin et al. , “Xnor-sram: In-memory computing sram macro for binary/ternary deep neural networks,” IEEE Journal of Solid-State Circuits, vol. 55, no. 6, pp. 1733–1743, 2020

  2. [9]

    Sram-based in-memory computing macro featuring voltage-mode accumulator and row-by-row adc for processing neural networks,

    J. Mu et al. , “Sram-based in-memory computing macro featuring voltage-mode accumulator and row-by-row adc for processing neural networks,” IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 69, no. 6, pp. 2412–2422, 2022

  3. [10]

    Efficient in-memory computing architecture based on crossbar arrays,

    B. Chen et al. , “Efficient in-memory computing architecture based on crossbar arrays,” in 2015 IEEE International Electron Devices Meeting (IEDM). IEEE, 2015, pp. 17–5

  4. [11]

    Four-layer 3d vertical rram integrated with finfet as a versatile computing unit for brain-inspired cognitive information processing,

    H. Li et al. , “Four-layer 3d vertical rram integrated with finfet as a versatile computing unit for brain-inspired cognitive information processing,” in 2016 VLSI. IEEE, 2016, pp. 1–2

  5. [12]

    Hermes core–a 14nm cmos and pcm-based in-memory compute core using an array of 300ps/lsb linearized cco- based adcs and local digital processing,

    R. Khaddam-Aljameh et al., “Hermes core–a 14nm cmos and pcm-based in-memory compute core using an array of 300ps/lsb linearized cco- based adcs and local digital processing,” in 2021 Symposium on VLSI Circuits. IEEE, 2021, pp. 1–2

  6. [13]

    A 64-core mixed-signal in-memory compute chip based on phase-change memory for deep neural network inference,

    M. Le Gallo et al. , “A 64-core mixed-signal in-memory compute chip based on phase-change memory for deep neural network inference,” Nature Electronics, vol. 6, no. 9, pp. 680–693, 2023

  7. [14]

    A crossbar array of magnetoresistive memory devices for in-memory computing,

    S. Jung et al. , “A crossbar array of magnetoresistive memory devices for in-memory computing,” Nature, vol. 601, pp. 211–216, 2022

  8. [15]

    A 22nm 4mb stt-mram data-encrypted near-memory computation macro with a 192gb/s read-and-decryption bandwidth and 25.1-55.1 tops/w 8b mac for ai operations,

    Y .-C. Chiu et al., “A 22nm 4mb stt-mram data-encrypted near-memory computation macro with a 192gb/s read-and-decryption bandwidth and 25.1-55.1 tops/w 8b mac for ai operations,” in 2022 ISSCC , vol. 65. IEEE, 2022, pp. 178–180

  9. [16]

    Challenges and trends indeveloping nonvolatile memory-enabled computing chips for intelligent edge devices,

    J.-M. Hung et al. , “Challenges and trends indeveloping nonvolatile memory-enabled computing chips for intelligent edge devices,” IEEE Transactions on Electron Devices, vol. 67, no. 4, pp. 1444–1453, 2020

  10. [17]

    Magnetoresistive circuits and systems: Embedded non-volatile memory to crossbar arrays,

    A. Agrawal et al. , “Magnetoresistive circuits and systems: Embedded non-volatile memory to crossbar arrays,” IEEE Transactions on Circuits and Systems I: Regular Papers , vol. 68, no. 6, pp. 2281–2294, 2021

  11. [18]

    In-memory computing in emerging memory technologies for machine learning: An overview,

    K. Roy et al., “In-memory computing in emerging memory technologies for machine learning: An overview,” in 2020 57th ACM/IEEE Design Automation Conference (DAC). IEEE, 2020, pp. 1–6

  12. [19]

    Domain-specific stt-mram-based in-memory computing: A survey,

    A. Yusuf et al., “Domain-specific stt-mram-based in-memory computing: A survey,” IEEE Access, 2024

  13. [20]

    Sram write-and performance-assist cells for reducing interconnect resistance effects increased with technology scaling,

    K. Cho et al. , “Sram write-and performance-assist cells for reducing interconnect resistance effects increased with technology scaling,” IEEE Journal of Solid-State Circuits , vol. 57, no. 4, pp. 1039–1048, 2022

  14. [21]

    Chip power-frequency scaling in 10/7nm node,

    P. Oldiges et al., “Chip power-frequency scaling in 10/7nm node,” IEEE Access, vol. 8, pp. 154 329–154 337, 2020

  15. [23]

    16.4 an 89tops/w and 16.3 tops/mm 2 all-digital sram-based full-precision cim macro in 22nm for machine-learning edge applications,

    Y .-D. Chih et al. , “16.4 an 89tops/w and 16.3 tops/mm 2 all-digital sram-based full-precision cim macro in 22nm for machine-learning edge applications,” in 2021 ISSCC, vol. 64. IEEE, 2021, pp. 252–254

  16. [24]

    Integrated photonics: bridging the gap between optics and electronics for enhancing information processing,

    L. M. Shaker et al. , “Integrated photonics: bridging the gap between optics and electronics for enhancing information processing,” Journal of Optics, pp. 1–13, 2023

  17. [25]

    Harnessing optical advantages in computing: a review of current and future trends,

    C. G. Kibebe et al. , “Harnessing optical advantages in computing: a review of current and future trends,” Frontiers in Physics , vol. 12, p. 1379051, 2024

  18. [26]

    Modulation and multiplexing in optical communications,

    P. J. Winzer, “Modulation and multiplexing in optical communications,” in Conference on Lasers and Electro-Optics. Optica Publishing Group, 2009, p. CTuL3

  19. [27]

    Photonics for artificial intelligence and neuromor- phic computing,

    B. J. Shastri et al., “Photonics for artificial intelligence and neuromor- phic computing,” Nature Photonics, vol. 15, no. 2, pp. 102–114, 2021

  20. [28]

    In-memory computing on a photonic platform,

    C. R ´ıos et al., “In-memory computing on a photonic platform,” Science advances, vol. 5, no. 2, p. eaau5759, 2019

  21. [29]

    Si microring resonator crossbar array for on-chip inference and training of the optical neural network,

    S. Ohno et al. , “Si microring resonator crossbar array for on-chip inference and training of the optical neural network,” Acs Photonics , vol. 9, no. 8, pp. 2614–2622, 2022

  22. [30]

    Parallel convolutional processing using an integrated photonic tensor core,

    J. Feldmann et al., “Parallel convolutional processing using an integrated photonic tensor core,” Nature, vol. 589, no. 7840, pp. 52–58, 2021

  23. [31]

    On-chip photonic synapses based on slot-ridge waveg- uides with pcms for in-memory computing,

    H. Zhang et al., “On-chip photonic synapses based on slot-ridge waveg- uides with pcms for in-memory computing,” IEEE Photonics Journal , vol. 13, no. 2, pp. 1–13, 2021

  24. [32]

    Deep learning with coherent nanophotonic circuits,

    Y . Shen et al. , “Deep learning with coherent nanophotonic circuits,” Nature photonics, vol. 11, no. 7, pp. 441–446, 2017

  25. [33]

    120 gops photonic tensor core in thin-film lithium niobate for inference and in situ training,

    Z. Lin et al., “120 gops photonic tensor core in thin-film lithium niobate for inference and in situ training,” Nature Communications , vol. 15, no. 1, p. 9081, 2024

  26. [34]

    A photonic deep neural network processor on a single chip with optically accelerated training,

    S. Bandyopadhyay et al., “A photonic deep neural network processor on a single chip with optically accelerated training,” in 2023 Conference on Lasers and Electro-Optics (CLEO) . IEEE, 2023, pp. 1–2

  27. [35]

    Neuromorphic photonic networks using silicon photonic weight banks,

    A. N. Tait et al. , “Neuromorphic photonic networks using silicon photonic weight banks,” Scientific reports, vol. 7, no. 1, p. 7430, 2017

  28. [36]

    Integrated all-photonic non-volatile multi-level memory,

    C. R ´ıos et al., “Integrated all-photonic non-volatile multi-level memory,” Nature photonics, vol. 9, no. 11, pp. 725–732, 2015

  29. [37]

    A 128 gb/s pam4 silicon microring modulator with integrated thermo-optic resonance tuning,

    J. Sun et al. , “A 128 gb/s pam4 silicon microring modulator with integrated thermo-optic resonance tuning,” Journal of Lightwave Tech- nology, vol. 37, no. 1, pp. 110–115, 2018

  30. [38]

    High-q and high finesse silicon microring resonator,

    J. Nijem et al. , “High-q and high finesse silicon microring resonator,” Optics Express, vol. 32, no. 5, pp. 7896–7906, 2024

  31. [39]

    A 4gs/s 4b flash adc in 0.18/spl mu/m cmos,

    S. Park et al. , “A 4gs/s 4b flash adc in 0.18/spl mu/m cmos,” in 2006 IEEE International Solid State Circuits Conference-Digest of Technical Papers. IEEE, 2006, pp. 2330–2339

  32. [40]

    A 6-b 1.6-gsample/s flash adc in 0.18-/spl mu/m cmos using averaging termination,

    P. C. Scholtens et al., “A 6-b 1.6-gsample/s flash adc in 0.18-/spl mu/m cmos using averaging termination,” IEEE JSSC , vol. 37, no. 12, pp. 1599–1609, 2002

  33. [41]

    A 750mw 24gs/s 12b time-interleaved adc for direct rf sampling in modern wireless systems,

    S. S. Kumar et al., “A 750mw 24gs/s 12b time-interleaved adc for direct rf sampling in modern wireless systems,” in 2023 ISSCC. IEEE, 2023, pp. 1–3

  34. [42]

    Background timing mismatch calibration techniques in high-speed time-interleaved adcs: A tutorial review,

    M. Guo et al. , “Background timing mismatch calibration techniques in high-speed time-interleaved adcs: A tutorial review,” IEEE TCAS-II: Express Briefs, vol. 69, no. 6, pp. 2564–2569, 2022

  35. [43]

    Interchannel mismatch calibration techniques for time-interleaved sar adcs,

    M. Bagheri et al. , “Interchannel mismatch calibration techniques for time-interleaved sar adcs,” IEEE Open Journal of Circuits and Systems , vol. 2, pp. 420–433, 2021

  36. [44]

    Design of energy-efficient cross-coupled dif- ferential photonic-sram (psram) bitcell for high-speed on-chip photonic memory and compute systems,

    M. A.-A. Kaiser et al. , “Design of energy-efficient cross-coupled dif- ferential photonic-sram (psram) bitcell for high-speed on-chip photonic memory and compute systems,” arXiv preprint arXiv:2503.19544, 2025

  37. [45]

    Scalable in-memory compute optical processor,

    S. Sunder et al. , “Scalable in-memory compute optical processor,” in Smart Photonic and Optoelectronic Integrated Circuits 2025, vol. 13370. SPIE, 2025, pp. 107–113

  38. [46]

    A laser-forwarded coherent transceiver in 45-nm soi cmos using monolithic microring resonators,

    N. Mehta et al. , “A laser-forwarded coherent transceiver in 45-nm soi cmos using monolithic microring resonators,” IEEE Journal of Solid- State Circuits, vol. 55, no. 4, pp. 1096–1107, 2020

  39. [47]

    High power single mode 1300-nm superlattice based vcsel: Impact of the buried tunnel junction diameter on perfor- mance,

    S. A. Blokhin et al. , “High power single mode 1300-nm superlattice based vcsel: Impact of the buried tunnel junction diameter on perfor- mance,” IEEE Journal of Quantum Electronics, vol. 58, no. 2, pp. 1–15, 2022

  40. [48]

    Scalable parallel photonic processing unit for various neural network accelerations,

    S. Du et al. , “Scalable parallel photonic processing unit for various neural network accelerations,” Photonics Research, vol. 12, no. 11, pp. 2691–2702, 2024

  41. [49]

    11 tops photonic convolutional accelerator for optical neural networks,

    X. Xu et al. , “11 tops photonic convolutional accelerator for optical neural networks,” Nature, vol. 589, no. 7840, pp. 44–51, 2021

  42. [50]

    In-memory photonic dot-product engine with electri- cally programmable weight banks,

    W. Zhou et al. , “In-memory photonic dot-product engine with electri- cally programmable weight banks,” Nature Communications , vol. 14, no. 1, p. 2887, 2023

  43. [51]

    Computing dimension for a reconfigurable photonic tensor processing core based on silicon photonics,

    H. Ouyang et al., “Computing dimension for a reconfigurable photonic tensor processing core based on silicon photonics,” Optics Express , vol. 32, no. 18, pp. 31 205–31 219, 2024

  44. [52]

    A 42-ghz tia in 28-nm cmos with less than 1.8% thd for optical coherent receivers,

    L. Aschei et al., “A 42-ghz tia in 28-nm cmos with less than 1.8% thd for optical coherent receivers,” IEEE Solid-State Circuits Letters, vol. 3, pp. 238–241, 2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.