Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity...

71
PHOTONICS RESEARCH GROUP 1 PHOTONICS RESEARCH GROUP Photonic reservoir computing using silicon chips Kristof Vandoorne, Pauline Mechet, Martin Fiers, Thomas Van Vaerenbergh, Bendix Schneider, Andrew Katumba, Floris Laporte, David Verstraeten, Benjamin Schrauwen, Joni Dambre and Peter Bienstman

Transcript of Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity...

Page 1: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 1

PHOTONICS RESEARCH GROUP

Photonic reservoir computing using silicon chips

Kristof Vandoorne, Pauline Mechet, Martin Fiers, Thomas Van Vaerenbergh, Bendix Schneider, Andrew Katumba, Floris Laporte, David Verstraeten, Benjamin Schrauwen, Joni Dambre and Peter Bienstman

Page 2: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 2

THE BLACK BOX

Page 3: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 3

What can this chip do?

Page 4: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 4

Several things!

• Do arbitrary boolean calculations with memory on a bitstream

• Recognise arbitrary 5-bit headers at 12.5 Gbps

• Perform speech recognition of isolated digits

• Does not consume any active power

• Easily upscalable to higher speeds

Page 5: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 5

How does it do it?

Using “Reservoir computing”, a brain-inspired technique to solve pattern recognition problems in a fast and power-efficient way

Page 6: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 6

WHAT IS RESERVOIR COMPUTING?

Page 7: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 7

What is reservoir computing?

• From field of machine learning (2002)

• Related to neural networks

• So far mainly in software

• Very successful:• Better than state-of-the-art digit recognition

• Speech recognition

• Robot control

• …

Page 8: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 8

Reservoir Readout

Reservoir computing

Don’t train the neural network, only train the linear readout

Page 9: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 9

reservoir state

readout

reservoir

nothing pebbles gritpebbles

grit

A hardware implementation…

Page 10: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 10

*

**

**

*

*

●●

● ●

x

y

z'

*

** *

***x'

y'

To higher order space

●●● ●

●●

Why does it work?

Page 11: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 11

PHOTONIC RESERVOIR COMPUTING

Page 12: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 12

Photonics

Photonic reservoirs

Page 13: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 13

• Faster• More power efficient• Richer dynamics in nodes• Light has a phase

Why photonics?

Page 14: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 14

OPTICAL AMPLIFIER NETWORKSThe very beginning…

Page 15: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 15

Looks like tanh, but positive signals only

Output

Use SOAs as neurons

Page 16: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 16

The gain in the SOA model is dependent on the input power and its own history

SOA model

Page 17: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 17

81 SOAs

Swirl topology

Page 18: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 18

5 female speakers, saying

10 times the same 10 digits,

ranging from zero to nine

Speech corpus

Page 19: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 19

• dynamics of light signal should be on time scale of SOA dynamics and chip delays

• convert 1 sec speech to 1 ns light signal

• 9 orders of magnitude upconversion

Time scales

Page 20: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 20

Word error rate

Page 21: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 21

Optimal delay

75 ps

187.5 ps

312.5 ps

Page 22: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 22

Absolute minimum

(phase controlled)Minimum

(phase averaged)

Reducing 2D plots to single number

Page 23: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 23

Controlling the phase offers clear advantage

Page 24: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 24

PASSIVE SILICON RESERVOIRSThe next step…

Page 25: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 25

What happens if you remove the SOAs?

Page 26: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 26

Passive Silicon reservoir

• silicon photonics: mature technology

• nodes become simple splitters/combiners

• non-linearity in readout suffices

• no need for amplifiers which consume power

• no longer limited by timescale of non-linearity

Vandoorne et al, Nature Comms, 5, 3541, 2014

Page 27: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 27

NL coming from the detector suffices!

Speech task: passive reservoirs (no amplifiers)

Page 28: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 28

16 node swirl network where 11 nodes could be measured from 1 input

Page 29: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 29

The input: 11136 bits modulated at 1531 nm with speeds between 125Mbit/s and 12.5Gbit/s

Page 30: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 30

First task: desired output should be the XOR of every bit with the previous bit.

Hard task in machine learning (non-linear!)

Page 31: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 31

Measurements and simulations for the XOR task correspond

Page 32: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 32

The XOR task can be solved at different speeds and different bit combinations

Page 33: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 33

Other Boolean tasks can be solved as well (with the same reservoir states)

Page 34: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 34

Header recognition

Page 35: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 35

Advantages

• Scalability: • Note that we spent a lot of effort to slow down the signal!

• Easily scalable to higher speeds by shortening the delays

• No active power consumption on chip

• Same generic chip can be used for• digital tasks (simulation confirmed by experiment)

• analog tasks (theory only, no suitable equipment)

Page 36: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP - CONFIDENTIAL 36

APPLICATIONS

Page 37: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 37

Telecom task: non-linear equalization of optical links

Page 38: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 38

Signal Equalization: Results….

Up to 200 km below FEC Limit

Metro Links

Equalization results with passive SOI chip

Page 39: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 39

Scaling this up

• PhResCo: recently started H2020 European project (KULeuven, IBM, UGent, Supelec, IHP)

• Integrated readout on chip:

out…

…in

reservoir readout

Page 40: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 40

First design: comparing 3 different technologies

2 x 9 Reservoir

BTO Test Structures

Si Readout BTO Readout

VO2 Readout

VO2 Test Structures

Page 41: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 41

Conclusions

Neuromorphic computing

is interesting new paradigm

for photonics information processing

Page 42: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 42

Flow cytometry

http://www.lifetechnologies.com

Page 43: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 43

Imec cell sorter

Integrated micro-fluidic

channels

On-chip high speed

cell sorting

On-chip Fast

high-resolution microscopy

Page 44: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 44

Computational complexity

➢ Complex convolution or sequence of 2D FFTs

➢ 512x512 pixels/image

➢ 1M cells/sec

➢ 48.8M Flops for reconstruction

➢ ~ 60 TFlops/sec including classification

http://www.top500.org/

# Site System Cores Perf.ormance[TF/sec]

Power [kW]

482 AutomotiveUnited States

IBM Flex System x240, Xeon E5-2670 8C 2.600GHz, Infiniband FDR IBM

8,336 157.7 181

Page 45: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 45

Algorithm Methods

classificationfeature

selectionnumerical

reconstruction

Lymphocytes

Monocytes

Granulocytes

Page 46: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 46

Real experimental data

k

1.39

1.37

1.34E

Direction of flow

Incident plane wave Scattered wave +

Direct wave

Detector plane

Microfluidic flow

chamber

Page 47: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 47

Neural network - pipeline

.

.

.

.

.

.

Input Layer Hidden Layer Output Layer

ANN: < 200 GigaFlop/sec !

Page 48: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 48

Three-part WBC classification Results

• Dataset of ~7500 non-purified WBC:

Granulocytes (59.8%),

Lymphocytes (34.6%),

Monocytes (5.6%)

• Use of 10 random folds for cross-

validating (CV) the results

• Adding noise to weights at fixed SNR

Page 49: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 49

Purified monocyte/granulocyte classification

Averaged classification results with increasing signal-to-noise ratio (from left to right: 30dB, 10 dB, 3 dB)

Class 1 = monocytesClass 2 = granulocytes

Page 50: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 50

Towards a hardware solution

.

.

.

.

.

.

Page 51: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 51

Conclusions

Neuromorphic computing

is interesting new paradigm

for photonics information processing

Page 52: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 52

EXCITABLE SILICON RINGS

Page 53: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 53

Building a photonic spiking neuron

= ?

Page 54: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 54

Research question

• People have seen excitability in photonics before, but never cascaded it on chip

• Can we cascade excitability on-chip using ring-resonator neurons?

Page 55: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 55

Thermo-optic effect causes redshift

Light circulation in ring resonance dip/peak

Heating of the ring redshiftT

ire

Page 56: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 56

Self-heating causes bistability

Light circulation in ring resonance dip/peak

Heating of the ring redshiftT

ire

Page 57: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 57

Free carriers cause blueshift

Light circulation in ring resonance dip/peak

Free carriers blueshift

ire

N

Page 58: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 58

Combination free carrier and thermal effect can cause self-pulsation

Light circulation in ring ~ ps

Cooling of the ring ~ 100 ns

Free carriers ~ ns

ire

N

T

1

2

3

Page 59: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 59

Simulations: bistability and self-pulsation

Q 6.25 104

25 pm62 pm

R 4 μm

dB3

r −

Page 60: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 60

Simulation: excitability

Wavelength and input power ‘near’ self-pulsation...

Page 61: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 61

Simulation: cascadability

Page 62: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 62

Experiment: self-pulsation

Page 63: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 63

Experiment: excitability

Pulses excited by external trigger signal:

Page 64: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 64

Experiment: cascadability

Page 65: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 65

Cascading rings = creating a delay line

t

t

t

t

tt

t

Page 66: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 66

Cascading rings = creating a delay line

t

t

t

t

tt

t

Max ~ 9-10 rings

t

Page 67: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 67

10 rings result in a ~200 ns delay of a 15-20 ns pulse

Page 68: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 68

10 rings result in a ~200 ns delay of a 15-20 ns pulse

Page 69: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 69

Making a loop => spike encoded memory/clock

t

t t

If delay > internal timescale neuron

=> Excitation loops through rings

Page 70: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 70

The concept works! (loop from ring 2-8)

Page 71: Photonic Reservoir Computing · P. HOTONICS. R. ESEARCH. G. ROUP. 44. Computational complexity Complex convolution or sequence of 2D FFTs 512x512 pixels/image 1M cells/sec 48.8M Flops

PHOTONICS RESEARCH GROUP 71

Conclusions

Neuromorphic computing

is interesting new paradigm

for photonics information processing