Content deleted Content added
m Open access bot: doi added to citation with #oabot. |
→Unconventional approaches: Added section on neuromorphic computing using integrated optics to update this section. |
||
Line 69:
The light will enter in Start node. It will be divided into two (sub)rays of smaller intensity. These two rays will arrive into the second node at moments ''a1'' and 0. Each of them will be divided into two subrays which
will arrive in the third node at moments 0, ''a1'', ''a2'' and ''a1 + a2''. These represents the all subsets of the set {''a1, a2''}. We expect fluctuations in the intensity of the signal at no more than four different moments. In the destination node we expect fluctuations at no more than 16 different moments (which are all the subsets of the given). If we have a fluctuation in the target moment ''B'', it means that we have a solution of the problem, otherwise there is no subset whose sum of elements equals ''B''. For the practical implementation we cannot have zero-length cables, thus all cables are increased with a small (fixed for all) value ''k'. In this case the solution is expected at moment ''B+n×k''.
=== On-Chip Photonic Tensor Cores ===
With increasing demands on graphical processing unit-based accelerator technologies, in the second decade of the 21st century, there has been an huge emphasis on the use of on-chip integrated optics to create photonics-based processors. The emergence of both deep learning neural networks based on phase modulation <ref>{{Cite journal |last=Shen |first=Yichen |last2=Harris |first2=Nicholas C. |last3=Skirlo |first3=Scott |last4=Prabhu |first4=Mihika |last5=Baehr-Jones |first5=Tom |last6=Hochberg |first6=Michael |last7=Sun |first7=Xin |last8=Zhao |first8=Shijie |last9=Larochelle |first9=Hugo |last10=Englund |first10=Dirk |last11=Soljačić |first11=Marin |date=2017-07 |title=Deep learning with coherent nanophotonic circuits |url=https://www.nature.com/articles/nphoton.2017.93 |journal=Nature Photonics |language=en |volume=11 |issue=7 |pages=441–446 |doi=10.1038/nphoton.2017.93 |issn=1749-4893}}</ref>, and more recently amplitude modulation using photonic memories <ref>{{Cite journal |last=Ríos |first=Carlos |last2=Youngblood |first2=Nathan |last3=Cheng |first3=Zengguang |last4=Le Gallo |first4=Manuel |last5=Pernice |first5=Wolfram H. P. |last6=Wright |first6=C. David |last7=Sebastian |first7=Abu |last8=Bhaskaran |first8=Harish |date=2019-02 |title=In-memory computing on a photonic platform |url=https://www.science.org/doi/10.1126/sciadv.aau5759 |journal=Science Advances |language=en |volume=5 |issue=2 |doi=10.1126/sciadv.aau5759 |issn=2375-2548 |pmc=PMC6377270 |pmid=30793028}}</ref> have created a new area of photonic technologies for neuromorphic computing <ref>{{Cite book |last=Prucnal |first=Paul R. |url=https://books.google.com/books?id=VbvODgAAQBAJ&newbks=0&hl=en |title=Neuromorphic Photonics |last2=Shastri |first2=Bhavin J. |date=2017-05-08 |publisher=CRC Press |isbn=978-1-4987-2524-8 |language=en}}</ref><ref>{{Cite journal |last=Shastri |first=Bhavin J. |last2=Tait |first2=Alexander N. |last3=Ferreira de Lima |first3=T. |last4=Pernice |first4=Wolfram H. P. |last5=Bhaskaran |first5=Harish |last6=Wright |first6=C. D. |last7=Prucnal |first7=Paul R. |date=2021-02 |title=Photonics for artificial intelligence and neuromorphic computing |url=https://www.nature.com/articles/s41566-020-00754-y |journal=Nature Photonics |language=en |volume=15 |issue=2 |pages=102–114 |doi=10.1038/s41566-020-00754-y |issn=1749-4893}}</ref>, leading to new photonic computing technologies, all on a chip such as the photonic tensor core <ref>{{Cite journal |last=Feldmann |first=J. |last2=Youngblood |first2=N. |last3=Karpov |first3=M. |last4=Gehring |first4=H. |last5=Li |first5=X. |last6=Stappers |first6=M. |last7=Le Gallo |first7=M. |last8=Fu |first8=X. |last9=Lukashchuk |first9=A. |last10=Raja |first10=A. S. |last11=Liu |first11=J. |last12=Wright |first12=C. D. |last13=Sebastian |first13=A. |last14=Kippenberg |first14=T. J. |last15=Pernice |first15=W. H. P. |date=2021-01 |title=Parallel convolutional processing using an integrated photonic tensor core |url=https://www.nature.com/articles/s41586-020-03070-1 |journal=Nature |language=en |volume=589 |issue=7840 |pages=52–58 |doi=10.1038/s41586-020-03070-1 |issn=1476-4687}}</ref>.
===Wavelength-based computing===
| |||