Multi-level Encoding and Decoding in a Wavelength-Multiplexed Photonic Tensor Processor

Guo, Zhimu; Márquez, Bicky A.; Filipovich, Matthew J.; Morison, Hugh; Shastri, Bhavin J.; Chrostowski, Lukas; Shekhar, Sudip; Prucnal, Paul R.

doi:10.1109/gfp51802.2021.9673887

Cited by 2 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, a wavelength of interest ( ) is selected, which is achieved by applying a voltage to the IRPH ( V IRPH ). For information encoding, a "discrete analog" encoding/decoding scheme is employed to implement direct value mapping for translating the digital number to an analog value 34,35 . More details about the "discrete analog" scheme can be found in "Dot product engine" Section.…”

Section: Intensity-modulated Microring Modulatorsmentioning

confidence: 99%

“…The power levels for the weight MRRs are obtained after subtracting each power level between drop (dashed red line) and through (dashed green line) ports. For the input MRR, the power levels are distributed between 0 to 15 µ W while for the weight MRR, due to the IL, the maximum transmitted power at the drop port can only reach 11.25 µ W; thus limiting 16 power levels between 11.25 to -15 µ W. We employ the "discrete analog" scheme to encode inputs and weights to distinct power levels of the input MRR and the weight MRR, respectively 35 . By normalizing each quantized power level (solid lines in Figure 9a) by the maximum measured output power of 15 µ W, a point-to-point mapping is realized by endowing voltage pairs with digital information, which can be expressed as: where, D is the correlated digital number, and I volt pair and I max are the voltage pair-applied transmission and the maximum transmission of the IM-MRM, respectively.…”

Section: Dot Product Calculationmentioning

confidence: 99%

See 1 more Smart Citation

Towards a high-density photonic tensor core enabled by intensity-modulated microrings and photonic wire bonding

Luan

Salmani

et al. 2023

Sci Rep

Self Cite

View full text Add to dashboard Cite

We propose a photonic processing unit for high-density analog computation using intensity-modulation-based microring modulators (IM-MRMs). The output signal at the fixed resonance wavelength is directly intensity modulated by changing the extinction ratio (ER) of the IM-MRM . Thanks to the intensity-modulated approach, the proposed photonic processing unit is less sensitive to the inter-channel crosstalk. Simulation results reveal that the proposed design offers a maximum of 17-fold increase in wavelength channel density compared to its wavelength-modulated counterpart. Therefore, a photonic tensor core of size 512 $$\times $$ × 512 can be realized by current foundry lines. A convolutional neural network (CNN) simulator with 6-bit precision is built for handwritten digit recognition task using the proposed modulator. Simulation results show an overall accuracy of 96.76%, when the wavelength channel spacing suffers a 3-dB power penalty. To experimentally validate the system, 1000 dot product operations are carried out with a 4-bit signed system on a co-packaged photonic chip, where optical and electrical I/Os are realized using photonic and electrical wire bonding techniques. Study of the measurement results show a mean squared error (MSE) of 3.09$$\times $$ × 10$$^{-3}$$ - 3 for dot product calculations. The proposed IM-MRM, therefore, renders the crosstalk issue tractable and provides a solution for the development of large-scale optical information processing systems with multiple wavelengths.

show abstract

Section: Intensity-modulated Microring Modulatorsmentioning

confidence: 99%

Section: Dot Product Calculationmentioning

confidence: 99%

Towards a high-density photonic tensor core enabled by intensity-modulated microrings and photonic wire bonding

Luan

Salmani

et al. 2023

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…First, a wavelength of interest (λ ) is selected, which is achieved by applying a voltage to the IRPH (V λ IRPH ). For information encoding, a "discrete analog" encoding/decoding scheme is employed to implement direct value mapping for translating the digital number to an analog value 34,35 . More details about the "discrete analog" scheme can be found in Section 3.…”

Section: Characterization and Controlmentioning

confidence: 99%

“…For the input MRR, the power levels are distributed between 0 to 15 µW while for the weight MRR, due to the IL, the maximum transmitted power at the drop port can only reach 11.25 µW; thus limiting 16 power levels between 11.25 to -15 µW. We employ the "discrete analog" scheme to encode inputs and weights to distinct power levels of the input MRR and the weight MRR, respectively 35 . By normalizing each quantized power level (solid lines in Figure 9(a)) by the maximum measured output power of 15 µW, a point-to-point mapping is realized by endowing voltage pairs with digital information, which can be expressed as:…”

Section: Dot Product Calculationmentioning

confidence: 99%

Towards a High-density Photonic Tensor Core Enabled by Intensity-modulated Microrings and Photonic Wire Bonding

Luan

Salmani

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

We propose a photonic processing unit for high-density analog computation using intensity-modulation-based microringmodulators (IM-MRMs). The output signal at the fixed resonance wavelength is directly intensity modulated by changing theextinction ratio (ER) of the IM-MRM. Thanks to the intensity-modulated approach, the proposed photonic processing unit isless sensitive to the inter-channel crosstalk. Simulation results reveal that the proposed design offers a maximum of 17-foldincrease in wavelength channel density compared to its wavelength-modulated counterpart. Therefore, a photonic tensor coreof size 512 × 512 can be realized by current foundry lines. A convolutional neural network (CNN) simulator with 6-bit precisionis built for handwritten digit recognition task using the proposed modulator. Simulation results show an overall accuracy of 96.76%, when the wavelength channel spacing suffers a 3-dB power penalty. To experimentally validate the system, 1000 dotproduct operations are carried out with a 4-bit signed system on a co-packaged photonic chip, where optical and electrical I/Osare realized using photonic and electrical wire bonding techniques. Study of the measurement results show a mean squarederror (MSE) of 3.09 × 10^−3 for dot product calculations. The proposed IM-MRM, therefore, renders the crosstalk issue tractableand provides a solution for the development of large-scale optical information processing systems with multiple wavelengths.

show abstract

Multi-level Encoding and Decoding in a Wavelength-Multiplexed Photonic Tensor Processor

Cited by 2 publications

References 9 publications

Towards a high-density photonic tensor core enabled by intensity-modulated microrings and photonic wire bonding

Towards a high-density photonic tensor core enabled by intensity-modulated microrings and photonic wire bonding

Towards a High-density Photonic Tensor Core Enabled by Intensity-modulated Microrings and Photonic Wire Bonding

Contact Info

Product

Resources

About