& cplSiteName &

Teradvance: A New Take on 40 Gig

Light Reading
News Analysis
Light Reading
9/14/2001

Earlier this year, Light Reading published a report that identified some of the challenges facing developers of 40 Gbit/s transmission systems (see 40-Gig Forecast).

They include the scarcity, cost and unproven performance of 40-gig components, big problems with dispersion and nonlinear effects and difficulties in using this technology over legacy fiber. Another challenge is also implicit in the report’s long list of companies developing systems in this field. The competition is extremely tough.

Enter Teradvance Communications, Inc., a startup that’s aiming to demonstrate that it’s got a couple of winning differences in this field.

First, it isn’t competing head-to-head with system vendors. It’s raised a relatively small amount of angel funding - $2 million – and is limiting its goals to developing 40-gig and 80-gig transmission technology and then selling it to a systems vendor. "A few people" have already shown some interest, the company says.

Teradvance is actually working on two versions of its technology – a pair of boxes (transmit and receive) for use in ultra-long haul, long haul and marine applications, and a much lower cost pair of chips for use in metro networks. The metro chips don't have the snazzy stuff needed for long distance transmissions.

The company expects to demonstrate a 40-gig proof-of-concept version of its technology later this month, and have a 80-gig prototype ready by November.

The second winning difference is that Teradvance believes that it's conquered many of the technical challenges facing 40-gig systems.

In particular, it’s found a way of splitting transmissions into four wavelengths each carrying 10 Gbit/s while using a single laser for all four wavelengths. This means that it can make use of well-proven, low-cost 10 Gbit/s components while avoiding the need for four expensive lasers, one for each wavelength, as would normally be the case.

That’s just for starters. Teradvance says its technology achieves a 4 decibel signal to noise (SNR) advantage over the normal return-to-zero modulation schemes used by other developers of 40-gig systems. This also minimizes the impact of non-linearities, polarization mode dispersion and chromatic dispersion, it adds.

The bottom line is that Teradvance's pair of boxes can carry as many as 80 40-gig channels in the C band over “any type of fiber”, and can do this over distances of up to 2,000 km, according to the company.

This eliminates requirements for forward error correction (FEC) or Raman amplication when using non-zero-dispersion-shifted-fiber (NZ-DSF). It also eliminates requirements for separate dispersion compensators, the company says.

In addition, Teradvance's SNR performance means that carriers aren’t faced with having to upgrade amplifiers and shorten their spacing, as they might well have to with alternative 40-gig technologies, according to the company.

Teradvance says that its closest competitor is probably OptiMight Communications Inc., which also makes a big thing out of achieving long transmission distances without resorting to Raman amplification (see Raman Risks Emerge and OptiMight Details Long-Haul Box).

In order to get an expert’s opinion on Teradvance’s technology, Light Reading invited Stuart Barnes, now with Atlas Venture, a venture capital company, to comment on a technology overview from the startup.

Until recently, Barnes was director of engineering at Ilotron, a startup that was developing, among other things, transmission systems before it ran into funding problems earlier this year (see Ilotron Hits Hard Times). Prior to that, Barnes worked for Alcatel SA (NYSE: ALA; Paris: CGEP:PA), developing transmission systems.

Barnes says that Teradvance’s spectral efficiency (how much carrying capacity can be squeezed out of fiber) looks better than Optimight’s. However, Alcatel, Tyco International Ltd. (NYSE: TYC) and “the Japanese” (probably NEC Corp.) (Nasdaq: NIPNY)) have even better spectral efficiency figures than Teradvance, according to Barnes. Another couple of un-named U.S. startups might also have comparable performance, he adds.

Barnes’ comments on Teradvance’s technology follow, together with responses from the startup’s technical team:

  • Barnes: Modulation scheme looks key to long reach. Other than polarisation multiplexing (to get in extra channels). They remain silent on the details here. Presumably this is key IP.
    Teradvance: We fully agree. The modulation scheme is key to our IP.
  • Barnes: PMD compensation is a must at 40G (but everyone’s in this space). Still if they have the chip sets in hand they are on the leading edge.
    Teradvance: We do not need chipsets that we do not have - so we are indeed on the leading edge. We do not require the same PMD compensation equipment that other systems need. This is a tremendous advantage for us since the compensation equipment required by other systems is not yet available or not yet cost effective.
  • Barnes: The 4dB Noise Figure claimed for amplifiers is complete tosh (English colloquialism for baloney) for a legacy system. NF's here are typically 6.
    Teradvance: We agree. However, our reach using amplifiers with a noise figure of 6dB is still well over 1,500 km on SMF. Regardless of the noise figure used, our system is always better than other systems by the same amount. That is, the performance delta is always the same.
  • Barnes: Have they tested the solution over the whole C band? Generally the greatest problems come with the external channels. If it just simulation, have they reconciled with some experimental data.
    Teradvance: The only effects that would kick in at the external channels are dispersion slope effects, which we can handle due to the higher tolerance to residual dispersion. We have indeed reconciled simulations and experiments.

    — Peter Heywood, Founding Editor, Light Reading
    http://www.lightreading.com
    (11)  | 
    Comment  | 
    Print  | 
  • Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
    Page 1 / 2   >   >>
    reconv
    reconv
    12/4/2012 | 7:51:02 PM
    re: Teradvance: A New Take on 40 Gig
    Is Teradvance another Optical FDM company (Kestrel, Centerpoint) with a solution focused on 40Gbps?

    dr
    dr
    12/4/2012 | 7:51:01 PM
    re: Teradvance: A New Take on 40 Gig
    The Teradvance technology is not optical FDM. It is more spectrally efficient than FDM, more power efficient, and it is more tolerant to PMD, chromatic dispersion and nonlinearities.
    spont
    spont
    12/4/2012 | 7:51:00 PM
    re: Teradvance: A New Take on 40 Gig
    sounds like they don't do 40G NRZ or RZ. that's for sure. 4*10G with a IPR waveform on each 10G is not really 40G. Or is it??
    sundevil
    sundevil
    12/4/2012 | 7:50:59 PM
    re: Teradvance: A New Take on 40 Gig
    Spont,

    I'm not familiar with the acronym IPR is this a type of coding format?
    Peter Heywood
    Peter Heywood
    12/4/2012 | 7:50:59 PM
    re: Teradvance: A New Take on 40 Gig
    Hello dr, do I know you?

    When I asked Teradvance about competition, it named Optimight and Kestrel, so even if this isn't optical FDM, it sounds as though it's equivalent to it.
    nobollox
    nobollox
    12/4/2012 | 7:50:56 PM
    re: Teradvance: A New Take on 40 Gig
    How would this solution differ from Kestrel's or Optimight? Isn't Optimight one of Wu-Fu Chen's companies?

    jbebop
    jbebop
    12/4/2012 | 7:50:54 PM
    re: Teradvance: A New Take on 40 Gig

    How many times are we going going to hear "we have this magical coding scheme that allows us to do 40G." Whatever, what a joke. $2M in financing is nothing. I dunno why LR gets all worked up about startups like this. Think about what $2M buys, not much, especially in this space. This company barely has enough financing to purchase the test equipment it needs let alone hire good people.

    "'A few people' have already shown some interest, the company says." Read this as a friend of one of the founders who works at a systems company and the LR reporter who asked.

    "The company expects to demonstrate a 40-gig proof-of-concept version of its technology later this month, and have a 80-gig prototype ready by November."

    Whatever, so what, how many university research labs have this going? Proof-of-concept, what a joke. LR how about making this news when they actually have something to sell.


    "This eliminates requirements for forward error correction (FEC) or Raman amplication when using non-zero-dispersion-shifted-fiber (NZ-DSF). It also eliminates requirements for separate dispersion compensators, the company says."

    Is this a joke...????? Why does LR give any creedance to something which is obviously just hype?


    Petabit
    Petabit
    12/4/2012 | 7:50:44 PM
    re: Teradvance: A New Take on 40 Gig
    I'm a little worried about Teradvance, their public relations image is starting to look a bit suspect, and if they are not careful they will head down the Silkroad route.

    Their problem is credibilty. If you come up with a new widget that will revolutionise the world, you better be in a position to demonstrate and explain how it works. Pons and Fleishmann (sp?) did a great job of demonstrating cold fusion, but could never explain how it worked to the satisfaction of the science community in general. Teradvance need to do a better job of explaining their technology, or they will lose credibility.

    Let's look at the description of the technology used in the article.

    "itGs found a way of splitting transmissions into four wavelengths each carrying 10 Gbit/s while using a single laser for all four wavelengths." The use of the word wavelengths is interesting - they are not described as channels. There are many ways to split wavelengths, you can use a GHz frequency signal to beat with the light (its called FDM, which Teradvance deny), you can use a second laser (which Teradvance deny), you can use scattering effects (which would buy you the wrong amount of frequency shift), or you could be using the wrong word. There are other ways to create multiple channels within one laser - multi level coding being the first one to spring to mind.

    "This means that it can make use of well-proven, low-cost 10 Gbit/s components while avoiding the need for four expensive lasers, one for each wavelength, as would normally be the case." The lasers are usually cheaper than the modulators at 10G, perhaps you've been confusing cost and price (like several other startups I could name).

    "Teradvance says its technology achieves a 4 decibel signal to noise (SNR) advantage over the normal return-to-zero modulation schemes used by other developers of 40-gig systems." Most 40G systems that have been proposed use a modulation format that is not too far from RZ, but offer SNR advantages over plain RZ. Not new, try reading some papers - OFC is a good start.

    "This also minimizes the impact of non-linearities, polarization mode dispersion and chromatic dispersion, it adds." 10G modulation is much more robust to dispersion than 40G. However as several people have found, inverse muxing doesn't buy you anything. Spliting a 40G signal into 4 10G streams, means that the dispersion tolerance BETWEEN THE 10G streams is as tight as the 40G signal - and you just gone and spread the data over a larger part of the spectrum (if they really are on separate wavelengths) - which makes the problem a whole lot worse. When Teradvance finally try and build some of this stuff (and $2 million won't get them very far towards that) they will be in for a nasty surprise.

    "The bottom line is that Teradvance's pair of boxes can carry as many as 80 40-gig channels in the C band over Gǣany type of fiberGǥ, and can do this over distances of up to 2,000 km, according to the company." 80 channels in the C band means that the channels are 50 GHz apart (if you want to use cheaper ITU spec components). By the time you include a mux and demux, the optical bandwidth of each channels drops to about 40 GHz. So Teradvance can't be using RZ or NRZ, nor can they be using solitons or SSB. They could be using polarisation muxing or multi-level coding, but that would not fit the desciption at the top of the article.

    And I can't let the 'any type of fibre' comment go. If Teradvance can operate on a Japanese link of 2000 km of DSF, with 80x 40G, then they are worth a lot more than $2 million. I don't think they can.

    "In addition, Teradvance's SNR performance means that carriers arenGt faced with having to upgrade amplifiers and shorten their spacing, as they might well have to with alternative 40-gig technologies, according to the company." You will have to upgrade the amplifiers, since their netowrk management sofware will not handle your modulation format. And since when has 40G needed shorter amplifier spacings? I seem to remember a paper at OFC that used 160km and 200km spacings for a 2000km 40G system.

    And finally: Mr Barnes, you do not need PMD compensation at 40G. None of the major manufacturers are using it.

    Summary:"ItGs raised a relatively small amount of angel funding - $2 million G and is limiting its goals to developing 40-gig and 80-gig transmission technology and then selling it to a systems vendor." Then it has an awful lot of work on its hands in order to pursuade the technical departments of those systems vendors, that their technology offers something better than the exisiting solutions.

    P.
    realguy
    realguy
    12/4/2012 | 7:50:37 PM
    re: Teradvance: A New Take on 40 Gig
    It may very well be true that this company offers no more better solution than any other hype driven firms out there, however I do not see why some people pour so much negativity on these guys. After all real technological breakthrough either happens by pure accident or by Edisonian minds with little regard for common wisdom.
    SPARKLE
    SPARKLE
    12/4/2012 | 7:50:34 PM
    re: Teradvance: A New Take on 40 Gig
    Hey, much of the content of this article reads exactly like an article I read early last week in THE OPTICAL KEYHOLE. What gives? I'm not getting timely stuff from you guys anymore.
    Page 1 / 2   >   >>
    Featured Video
    Upcoming Live Events
    October 22, 2019, Los Angeles, CA
    November 5, 2019, London, England
    November 7, 2019, London, UK
    November 14, 2019, Maritim Hotel, Berlin
    December 3-5, 2019, Vienna, Austria
    December 3, 2019, New York, New York
    March 16-18, 2020, Embassy Suites, Denver, Colorado
    May 18-20, 2020, Irving Convention Center, Dallas, TX
    All Upcoming Live Events
    Partner Perspectives - content from our sponsors
    Sports Venues: Where 5G Brings a Truly Immersive Experience
    By Peter Linder, 5G Evangelist, North America, Ericsson
    Multiband Microwave Provides High Capacity & High Reliability for 5G Transport
    By Don Frey, Principal Analyst, Transport & Routing, Ovum
    All Partner Perspectives