Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Monday January 10 2022, @05:27PM   Printer-friendly
from the can-you-hear-me-now? dept.

A silicon photonic-electronic neural network that could enhance submarine transmission systems:

We are currently witnessing an explosion of network traffic. Numerous emerging services and applications, such as cloud services, video streaming platforms and the Internet of Things (IOT), are further increasing the demand for high-capacity communications. Optical communication systems, technologies that transfer information optically using fibers, are the backbone of today's communication networks of fixed-line, wireless infrastructure and data centers.

Over the past decade, the growth of the internet was enabled by a technique known as digital signal processing (DSP), which can help to reduce transmission distortions. However, DSP is currently implemented using CMOS integrated circuits (ICs), thus it relies heavily on Moore's Law, which has approached its limits in terms of power dissipation, density and feasible engineering solutions.

As a result, distortions caused by a phenomenon known as fiber nonlinearity cannot be compensated by DSP, as this would require too much computation power and resources. Fiber nonlinearities remain the major limiting effect on long-distance transmission systems.

Researchers at Princeton Lightwave Lab and NEC Laboratory America have recently created a new neural network hardware that could help to overcome this limitation, compensating for the adverse effects of fiber nonlinearity. This neural network, presented in a paper published in Nature Electronics, is run on a silicon-based photonic-electronic system composing of a few neurons, which can, in principle, outperform commercial DSP chips in throughput, latency and energy use."

"The research on 'neuromorphic photonics' at Princeton began with a discovery by our supervisor, Prof. Paul Prucnal, and neuroscientist David Rosenbluth," Chaoran Huang, one of the researchers who carried out the study, told Tech Xplore. "These two researchers found that photonic devices and biological neurons are governed by identical differential equations, yet 'photonic neurons' have a time scale of roughly picosecond to nanosecond whereas biological neurons have a time scale of roughly one millisecond."

Highly recommend reading the — long and detailed — linked article.

Journal Reference:
Chaoran Huang, Shinsuke Fujisawa, Thomas Ferreira de Lima, et al. A silicon photonic–electronic neural network for fibre nonlinearity compensation, Nature Electronics (DOI: 10.1038/s41928-021-00661-2)
Broadcast and Weight: An Integrated Network For Scalable Photonic Spike Processing, (DOI: 10.1109/JLT.2014.2345652)


Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: -1, Flamebait) by Anonymous Coward on Monday January 10 2022, @05:53PM (2 children)

    by Anonymous Coward on Monday January 10 2022, @05:53PM (#1211553)

    >> the growth of the internet was enabled by a technique known as digital signal processing

    I thought the growth was enabled by AT&T and Spectrum ass-raping their customers by overcharging for services.

    • (Score: 0) by Anonymous Coward on Monday January 10 2022, @05:56PM (1 child)

      by Anonymous Coward on Monday January 10 2022, @05:56PM (#1211556)

      ... while collecting huge government handouts.

      • (Score: 2) by JoeMerchant on Monday January 10 2022, @06:51PM

        by JoeMerchant (3937) on Monday January 10 2022, @06:51PM (#1211576)

        As I recall, the .com (as in Pets.com, MySpace.com, etc.com) bust left Trillions of dollars wondering where to reinvest themselves, and fiber networks got a big chunk at that time. The abused government handouts were supposed to complete the last-mile connections that the ex-.com investors neglected, but why actually deliver any promised products of a government handout when your industry is "too big to fail"?

        --
        🌻🌻 [google.com]
  • (Score: 2) by JoeMerchant on Monday January 10 2022, @06:49PM (3 children)

    by JoeMerchant (3937) on Monday January 10 2022, @06:49PM (#1211575)

    If a couple of neurons can outperform a DSP, I'd wager that the DSP is "doing it wrong" and should be able to re-implement what's going on in those neurons with capacity to spare. (Assuming, a lot, of course.)

    --
    🌻🌻 [google.com]
    • (Score: 0) by Anonymous Coward on Monday January 10 2022, @10:22PM

      by Anonymous Coward on Monday January 10 2022, @10:22PM (#1211624)

      Is it like, accurate but contaminated signal vs lossy signal. And the latter wins by being smaller/faster?

    • (Score: 2) by crafoo on Tuesday January 11 2022, @12:38PM (1 child)

      by crafoo (6639) on Tuesday January 11 2022, @12:38PM (#1211741)

      Why do you believe that? Do you usually expect digital gates and flip-flops to beat out purpose-built equipment that operates with optical transmission components?

      • (Score: 2) by JoeMerchant on Tuesday January 11 2022, @02:21PM

        by JoeMerchant (3937) on Tuesday January 11 2022, @02:21PM (#1211761)

        When the signals are ultimately digital? Yes.

        Are they transmitting a million+ multiplexed voice signals in pure analog over a fiber with these neurons? If so, color me impressed, and also skeptical that analog voice transmission is where the money is, these days.

        --
        🌻🌻 [google.com]
  • (Score: 0) by Anonymous Coward on Tuesday January 11 2022, @02:13PM (1 child)

    by Anonymous Coward on Tuesday January 11 2022, @02:13PM (#1211757)

    pretty cool. if only now they could apply something like this while -or- during fiber making process ... better fiber?

    • (Score: 0) by Anonymous Coward on Tuesday January 11 2022, @02:20PM

      by Anonymous Coward on Tuesday January 11 2022, @02:20PM (#1211759)

      at one end you put in "the bible" and out comes "the koran". to test just send back what you received ... if you get the bible, all is good :)

(1)