Innovating the network edge with 100ZR QSFP28: The next frontier in coherent optics

0:01 hello everyone I'm Sean Buckley editor-in-chief of LightWave and BTR and it's my pleasure to welcome you to today's LightWave and BTR webcast where

0:08 we'll discuss innovating the network edge with 100 ZR qsfp28 the next

0:14 Frontier in coherent Optics in this webinar you'll get to hear from experts from juniper exfo and Precision Optical

0:21 Technologies who are teaming up to showcase the new 100z qsfp pluggable

0:27 coherent technology exploring its foundational cap abilities in diverse use cases in Access

0:33 aggregation and service edge networks participants will gain valuable insights on how to validate and deploy 100 ZR

0:40 qsfp28 coherent transceivers both in the labs and of course in the field um first

0:46 off I just want to thank our sponsors for making this webinar happen today beldin Precision Optics xfo and Juniper

0:52 Networks uh today's speakers will feature key experts in Optical routing and testing uh Annabelle alacorn product

0:59 manager for xfo uh we've got Paul touris compta principal Solutions engineer for

1:04 Precision Optical networks in beldin and AOG Den duri product manager for Juniper

1:11 Networks but before I hand things off to our speakers let me quickly review our webcast uh set up for those of you who

1:17 may be joining us for the first time or for the first time in a long time I'll start with you ask a question box you

1:22 should see near the slide area I invite you to feel free to ask questions at any time during the event today submit a

1:27 question you just click on that box type in your question question and hit the send button we'll address your questions after everyone has finished their

1:34 presentations now in the event they've run into a technical problem during the presentation you can out call for help

1:39 via the Q&A tool just describe your problem hit the question button and Milan who is today's webcast technician

1:46 will be happy to help you out next if you know of someone that might benefit from seeing this webcast or you just want to revisit the presentation

1:52 yourself this event will be available on demand from our webcast archive for the next several months we'll send you an

1:57 email with a link to the archive sometime in the next day or so so that you should feel free to forward to your co-workers and the archive version of

2:04 the webcast will also be available from our webcast AR archive section of our website which you can reach via our

2:10 homepage at www.l LightWave online.com if you'd like you can also be able to download a copy

2:17 of today's slides via the events resources tab you should see in your screens right below the ask a question

2:23 box just click the tab to reveal the slide files and with that uh let's look at our uh know this exciting new um

2:31 technology developed on on coherent Optics Trends I'm going to hand it off to Paul Paul the floor is yours uh take

2:37 to get us started thank you all

2:43 right thank you Sean for the intro um this webinar has been a great

2:48 collaborative effort with xfo and Juniper in my section I'll I'll focus on

2:54 the optic sides of the 100 ZR transceiver solution

3:01 so just to get start some quick background on how we got to this point different Technologies in use for

3:07 optical transceivers um historically IMDd and tensi modulated direct detect

3:13 Optics has been used across the board typically for lower um lower data rates one gig 100

3:21 Gig shortly reach applications pan for modulation IMDd Optics started getting

3:27 into uh the networks as higher data rates came into play and then historically coherent was reserved more

3:35 for the Long Reach applications uh Long Haul higher data

3:40 rates but as we've seen over the last few years that coherent technology has been getting into the small for factors

3:47 cfp qpd osfp and qsfp28

3:52 too now if we separate this in a couple of these blogs um IMDd coherent

3:58 technology in IMDd uh uh Optics which just doing basically an amplitude modulation of the signal

4:06 nrz that means just ones and zeros two different signal levels represent ones and zeros P four there is four different

4:13 states that represented by two different bit two bits that means effectively

4:18 we're doubling the bit rate um of of the signals that transmit and then in

4:24 coherent technology we're not only doing amplitude of the modulation we're doing also modulation of the phas with q or

4:32 qpsk modulation techniques and also there's new techniques such as dual

4:37 polarization all these techniques all these new algorithms require the development of new technologies such as

4:44 a coherent DSP that is able to perform um this data

4:51 processing all these technology development is intended to overcome some

4:57 of the limitations intrinsic to fiber optical Communications um one of the

5:03 main causes of limitation is the attenuation oan versus cban typically

5:09 DDM we're using cban is the lowest attenuation DB per kilometer that we can

5:14 achieve however in IMDd Optics using uh Optics that using the Cen you have

5:21 another big limiting factor that's chromatic dispersion and in this limiting factor really uh goes higher

5:29 the higher the day rate going from 10 gig to 25 gig DDM on MDD Optics or from

5:34 10 gig to 100 Gig it really limits the reach you can have with this kind of

5:40 Optics and then finally Optical signal to noise Rao just because we have a

5:46 signal that is being transmitted and received doesn't mean the signal is good enough for the receiver to understand it

5:53 so in order to overcome these challenges as we talked about there's been a development on the coherent optic space

5:59 on the coherent DSP space in order to compensate chromatic dispersion in order

6:05 to handle these modulation techniques and this Optics

6:11 transmission which brings us to the 100 ZR qsp 28 solution what makes this

6:18 solution is the incredibly powerful that we can put all this development on Cohen

6:25 Optics into a QP 28 F Factor we've seen it in cfp cfp to qpd OSP form factors

6:32 but the fact that we can put all this technology in a qsp 28 form factor which

6:37 by far has been the most deployed uh form factor for 100 Gig connectivity in in networks is is truly asstounding

6:45 this transceiver is a full tunable dwm optic with coherent optic coherent DSP

6:53 that allows the modulation of the signal with dual priorization dqpsk it has also in build f for a

6:59 correction as well as in internal chromatic dispersion

7:06 compensation this transceiver this product can be used in a wide range of

7:11 of applications for uh network service providers whether it's for unamplified grid links for higher than 100

7:18 kilometers point to point whether it's also for unamplified dwdm links with different reaches depending on the MOX

7:25 demo configuration and also for pointto Point Amplified d links when paired

7:31 together with an optical line system which reaches up to 300

7:37 kilomet now one thing that's is important to to remind here is that the

7:43 100 G usp28 is not so much just a single transceiver but a whole product line uh

7:50 of transceivers that can be customized to fit Network specific requirements so

7:56 just starting from the from the top whether this transer needs to be implemented in some of the latest and

8:03 newest q28 host that can support sis or whether this needs to be implemented in

8:09 some of the Legacy holes that are still in the network that only support sfff 8636 there's going to be a solution for

8:17 that also in terms of TX output power we're going to have like two different options whether it's low TX input highx

8:24 input that's specially relevant when going to the line system we've seen it happen in 400r plus with different

8:30 interations depending on the TX output power data rate options for 100 Gig

8:36 Ethernet or ot4 also important uh where these Optics are going to be placed

8:41 where it's in controlled environment there's going to be an option for cm and also it 10 variation for a wider range

8:48 of temperatures there's going to be also option for chromatic dispersion setting to extend that range we talked up to 300

8:55 kilometers via line system that's something the coherent DSP can can can be

9:00 enabled also for applications that require uh tunable Optical filter to be part of the TOA um so in colorless RM

9:09 applications for example that could be an option that can be added to this transceiver and finally also different

9:16 tuning modes of operation whether it's manual tuning via CLI or externally or

9:21 whether it's a to tune all these are product customizations that can

9:28 occur now if you saw in the previous slide we had different symbols that represent power consumption heat

9:34 dissipation one of the main challenges one of the main deployment considerations of these transceivers is

9:39 that they do consume quite a bit of power just to put it in perspective q28

9:45 100 G R4 most people is aware of it can consume up to 3.5 4.5 watt depending on

9:51 the generation of this optic a QP 2800 ZR depending on the product customization we're going to be talking

9:58 5.5 up to 6.5 watt so the implementation the installation of this optic is going

10:03 to be very much host platform depend dependent some of the new 400 gig host capables or 100 Gig host capable they

10:11 can handle even up to 15 20 watts per Port that should be no problem Legacy

10:17 host on the other hand maybe the QP 28 ports can only handle up to five five or

10:22 six Watts so certain products will be able to be used certain other products not there's going to be some

10:28 considerations when when deploying these Optics the amount of Optics we could plug depending on the host the more

10:34 power all these data processing we're talking about the coherent DSP and these Optics the more power the more heat is

10:40 generated so that's also something that needs to be taken into account when installing these transceivers in a

10:46 switch with several qsfp28 ports and last but not least the

10:52 compliance and compatibility piece these 100 ZR qcp 20 Optics do have the option

10:58 uh to be managed uh by simis however a lot of the network footprint installed

11:04 nowadays with qsa 2800 gig does not support the the Sim capability so we're

11:10 going to see as Juniper talk here uh today in this webinar there's new switches there's new software versions

11:17 that will gradually support 100 Gig ZR simis alternatively for customers that

11:22 need the implementation of q00 ZR nowadays with some of the Legacy

11:28 hostes that for different reasons can be upgraded to different software versions The 100gr sff with external tunable

11:35 capability will be an option for customers that need it right now and for that solution is one of the

11:43 expertise that prision has developed over the years over the last decade whether it was one gig tunables 10 gig

11:49 25 gig or now 100 Gig tunables we have our transer management modules the DN

11:55 100 QD in this example together with 100 transceiver you can plug it in the

12:01 module use our transceiver tool um to tune this transceiver to the desired

12:07 frequency the desired wavelength for your application and then plug it back

12:13 into your host router and make it work thank you

12:29 okay thank you Paul so yeah I think I think Paul did a fantastic job providing

12:35 an overview of the 100 Gig ZR transceivers themselves and kind of the coherent technology that enables it but

12:41 you know when you actually you know go up the solution stack from the plug bles to the overall Network architecture a

12:48 lot of the industry excitement around 100 Gig ZR is actually tied to this broader transition happening in the

12:53 market towards what we call IP over dwdm so I just want to spend a few minutes to kind of demist the architecture which

13:00 has gained a lot of traction and a lot of Mind share over the past few years so the diagram on the top half of this

13:05 slide represents a traditional approach to dwdm right you can see there's kind of a clear separation between your IP

13:11 layer where your routers live and the optical transport layer where you've got an optical transponder interconnecting

13:17 to an optical line system and when you look at this diagram left to right that interconnection between the two layers

13:24 is happening through a dedicated Optical transponder which has the the role of

13:29 taking those incoming gray uh or non-tunable typically client signals coming from the router uh and converting

13:36 them into colored coherent wavelengths tunable wavelengths such that they can be multiplexed onto a composite signal

13:42 Amplified and sent across a single fiber pair but you know there's nothing inherently wrong with this architecture

13:48 but fundamentally when you have a transponder demarcating the separation between the two layers you end up with

13:54 more physical devices in your network that you have to purchase install and manage uh in complete ilos which impacts

14:00 your network visibility it also inherently increases your capex spend and your ongoing Opex from a power

14:06 consumption and a space perspective overall your operational complexity of running a multi-layer network is high in

14:13 contrast the bottom diagram represents IP over dwdm uh you may also hear the terms you know packet Optical

14:19 integration converged Optical routing all of these terms refer to the same architectural change which is

14:25 integrating the dwdm transponder functionality into the IP router so

14:30 today that integration is made possible in a scalable and footprint optimized way through the use of tunable coherent

14:37 Optics like 400 gig and now 100 Gig ZR directly into router ports and you know

14:43 as Paul showed us the integration of advanced coherent Digital Signal processing into small form factors like

14:50 qsfp28 is allowing operators to overcome some of the key performance limitations

14:55 of you know IMDd based client Optics that we traditionally use in inside router rep reports so now when you look

15:01 at that bottom diagram left to right you can see that the coherent pluga bles have eliminated the need for a dedicated

15:07 transponder which yields a flat single layer kind of Ip Optical architecture with fewer power hungry and space

15:14 occupying Hardware devices in your topology um that you have to deploy and manage and not to mention you have a

15:20 rich IP service layer with vastly improved multi-layer visibility right now with the coherent pluga bles sitting

15:27 directly inside the router port the router has access to uh live Telemetry data uh about the health of the optical

15:34 Network right which can actually enable a wide array of interesting multi-layer management and automation uh use

15:43 cases now you know the entire promise of converging um the optical functionality

15:49 at the IP layer only works if your IP layer is designed to handle kind of the hardware and software requirements of

15:56 coherent Optics and you know at face value might be thinking you know 100 Gig ZR 400 gig ZR it's just another plugable

16:02 right um you know how hard could it be to just plug it into my router and get going you know I think the right way to think about it is that 100 Gig ZR is not

16:09 just another plugable it's really a building block for a high performance IP Optical solution and as a as a routing

16:16 vendor that means that if we want to deliver a plug-in-play experience for our customers there are really unique system design characteristics that we

16:23 need to consider um first is having purpose-built Hardware you know although as showed you advances in the coherent

16:31 DSP domain have brought down the power consumption of these plugs to you know a five to 6 and a half watt envelope

16:37 that's still roughly a what 60 to 70% increase when you compare it to the typical client Optics we use in routers

16:43 so as a matter of just Baseline feasibility the router has to have adequate power provision to the 100 Gig

16:49 port for this solution to be viable but power is just one variable right the more interesting challenge as we found

16:55 is the resulting thermal management burden that falls on the and and thermals is really not just a

17:01 linear function of power right when we at Juniper are building new systems we're placing special emphasis on

17:07 designing the system not only to have adequate air flow to cool the Optics and the physical hotspots you know on the

17:13 plugable itself but also to account for the heat contribution that the Optics has on the system for example on onboard

17:21 components like the routing Asic or the switch fabric which also have a propensity to get hot uh and in order in

17:27 order to actually do this successfully we're incorporating a lot of deliberate design choices that depart from how

17:34 Legacy routers used to be built right uh and one example of this is is the use of belly-to-belly cages um so traditionally

17:42 you'll find that routers used stacked qsfp cages um sometimes even without

17:47 dedicated heat sinks which creates cooling challenges especially for Optics that sit in the bottom row right uh so

17:53 with belly to belly cages meaning the top row qsfp cages are upright the bottom row is upside down you actually

18:00 end up with a more even distribution of heat uh which is an enabler for unrestricted support for coherent Optics

18:06 across the entire router or line cart right we often pair belly-to-belly cages with dedicated heat sinks to make that

18:12 happen uh another interesting kind of design choice that you'll find on many of Juniper's 100 Gig routers uh that sit

18:19 at the edge of the network is the use of qsfp DDD cages which is the mechanical form factor designed uh originally for

18:25 400 gig but we use those for dedicated 100 Gig Network ports um because qsfp

18:31 DDD is mechanically backwards compatible to qsfp28 this allows us to build a dedicated 100 Gig Port that takes

18:38 advantage of the power thermal and heat syncing Innovations introduced in the qsfp DD generation for higher speed

18:45 Optics but we use it to give us more thermal and power margin for 100 Gig Optics um as a great example of how this

18:52 all kind of comes together you can see on this slide we're showing a juniper ACX 7100 router this is a 1ru

18:59 4.8 terbit Metro router uh that's actually able to support a full density of 100 Gig and 400 gig zrzr plus Optics

19:07 with no restrictions on power or thermals the second important quality of a system that's you know ready for 100

19:14 Gig ZR is having a networking OS that's designed to meet the needs of a multi-layer converged Network um first

19:22 by way of the more complex and capable coherent dsps that sit inside the plug bles 100 Gig ZR has you know more

19:29 operating modes more parameters that need to be configured and monitored compared to a relatively simple client

19:35 inter client interface like a like a 100 Gig LR4 right uh this is why Juniper is a very strong Advocate and a contributor

19:41 to Industry standards like seus and coherent Semus which is you know really it's short for the common management

19:47 interface specification which are industry standards that are attempting to kind of take all of this complexity

19:53 and put it into a standard host to module interface so whether it be uh the ability to configure the wavelength of a

20:00 ZR optic uh monitor all of the key performance parameters uh you know navigating and selecting through the

20:06 different you know FEC modulation channelization modes advertised by the DSP to the host the underlying logic in

20:14 our Network OS is compliant to seus with the intent of enabling a plug-and-play

20:19 you know robust experience for our customers second is again intelligent thermal management right we touched on

20:26 some of the fundamental Hardware design attributes associated with Cooling and providing power to coherent Optics but

20:32 to actually cool the Optics while still optimizing for things like fan speed rotations and power consumption of the

20:38 system the host OS really needs to start thinking about coherent Optics not as an

20:43 external plugable but as an extension of the system right it's another sensor for the OS to pull when making environmental

20:51 decisions and flagging alarms to the users because you know with coherent plug bles routers may very well

20:57 encounter situations where the thermally limiting component is not the Asic it's not the fabric it's actually the optic

21:02 that's heating up faster it's the optic itself right so making those thermal management decisions uh needs to be kind

21:09 of part of the network OS environmental monitoring logic uh and lastly a support

21:14 for open apis and data models to kind of piece the solution together uh whether it be you know thirdparty Optical or

21:20 multi-layer controllers you know te Telemetry visualization tools or automation workflows with coherent

21:27 Optics sitting in the router the key building block to take advantage of having this multi-layer integration is

21:33 having the router expose configuration and Telemetry knobs to be consumed uh in

21:39 a vendor agnostic way right and this is why in the junos operating system we've made substantial Investments to expand

21:45 support for things like open config uh specific to the management of coherent transceivers through you know with bi

21:52 vendor agnostic ecosystem so now that we hopefully

21:59 understand what IP over dwdm is you know why people care about it and some of the key system design considerations I want

22:05 to switch gears and talk a little bit about the use case Matrix and what sort of network deployments we're expecting

22:11 to see a lot of traction with the 100 Gig ZR so rewinding back a few years right the momentum around this

22:17 architecture of Ip over dwdm didn't start with 100 Gig right it actually started with 400 gig ZR uh which is

22:24 which was for the first time a fully standardized uh and power and for form factor optimized coherent plugable right

22:31 and the 400 gig ZR was very quickly adopted particularly by Cloud providers and hyperscalers for relatively simple

22:37 data center interconnect use cases uh a year or two later we began to see the 400 gig ZR plus uh including modules

22:44 with higher zero dbm transmit output Powers finally extending the architecture to the service provider

22:50 world right primarily in the Metro and the edge but also some Regional multipan topologies roughly 500 to 1,000

22:57 kilometers and now in 2025 we're at a very interesting inflection point because we're starting to see both the

23:03 scale up to 800 gig and the scale down to 100 Gig in the pluggable coherent

23:08 world right so 800 gig zrzr plus is now a reality and in fact we at Juniper

23:14 recently put out a press release with one of our customers Alisa showcasing the world's first production 800 gig ZR

23:19 deployment right so it's very much real and very much in the field with 800 gig you not only get the next generation of

23:25 DCI with lower cost per cost and Power for bit but you unlock the ability to service kind of the the most performance

23:33 demanding applications like longfall transport but simultaneously in the industry what you see is the scale down

23:39 right getting a power and form factor optimized 100 Gig solution is also now a reality right um we're we're finally

23:46 unlocking new scalable IP over dwdm architectures on the lower end the edge and access networks so now that you've

23:54 got a cost and power optimized qsfp28 pluggable that that's able to sit in a

23:59 dedicated router 100 Gig Port you're able to bypass transponders and close up to 300 km links which was previously

24:06 unheard of in this footprint so where will we see the Demand right uh when you look at access networks today a lot of

24:13 Downstream links are being upgraded from 1 gig to 10 gig or in some cases even to 25 gig or 50 gig particularly in Mobile

24:21 you know back front mid Hall architectures which means increasingly 10 gig dwdm which is deployed in volume

24:27 today in the field May no longer be sufficient uh and operators will seek out 100 gig uplinks even if it's

24:34 initially just a few to their aggregation sites uh or even from the aggregation site to the network Edge

24:39 right and whether it be due to fiber quality or distance or wanting the dwdm capability having those 100 Gig uplinks

24:47 be coherent 100 Gig ZR is starting to make a lot of sense for a lot of customers um we also have customers with

24:54 smaller scale kind of DCI or Enterprise Edge deployments where 400 gig ZR never

24:59 really quite made sense because it was too much capacity or too much

25:09 power in the right footprint for the

25:17 job customer deployments will be with you know probably commercial temperature

25:23 and uh low low minus 8 dbm TX output power modules but now that the key

25:29 Technology Building Block exists uh and more and more you know DSP and Optical module vendors will enter the market

25:36 there are additional flavors of 100 Gig ZR that will emerge over the next year or so uh the first one uh which will be

25:43 which is already readily available is the industrial temperature 100 Gig ZR for temperature hardened applications so

25:49 this is particularly relevant for uncontrolled or outdoor deployments where your routers and Optics could be

25:55 exposed to temperatures from you know- 40° C through 85 deges CI uh here you

26:01 can see on the left hand side kind of a simplified diagram of a radio Access Network right using uh for example a

26:08 temperature hardened router like the Juniper ACX 7024 in a kind of 5G front

26:13 back mid Hall roll and a temperature hardened access router like the Juniper ACX 7348 in the aggregation role so we

26:21 have customers building mobile access networks who are planning to deploy 10 gig or 25 gig industrial temperature

26:27 Optics for Downstream connectivity to the you know 5G cell towers and then utilize the 100 Gig ZR industrial

26:34 temperature as an Uplink back to the aggregation router the second variant of 100 Gig Z that will probably drive a lot

26:40 of interest is with higher TX output Powers right of zero dbm particularly and I think there's two primary drivers

26:47 for this right first is extending your reach performance over dark fiber in an unamplified link the the amount of reach

26:54 you achieve is limited by loss right whether that be your you know patch panel connector losses uh insertion loss

27:00 as Paul said going through a mux uh or even just your fiber attenuation your fiber quality right so starting with a

27:06 higher transmit output power gives you additional link loss compensation and for some of our customers this is great

27:12 because they'll be able to push you know 120 kilometers without any external amplifiers which further simplifies

27:18 their deployment for others having a higher TX output power is kind of a

27:24 baseline requirement needed to compensate for older poor quality fiber they have deployed uh with no immediate

27:30 plans to kind of install new fiber right um the other use Cas is for interop over rodm infrastructures which typically

27:37 expect a per wavelength uh TX power of at least zero dbm uh for example we have

27:42 smaller service provider customers who have incumbent Metro dwdm ring topologies with multiple add drops

27:48 typically 100 to 150 kilometers of distance for them 400 gig ZR plus was

27:54 way too much capacity and again not in the right form factor for the 100 gig optimized routers that they have already

28:00 deployed in the field in those roles so the 100 Gig ZR zbm makes a lot of sense and it'll enable them to move away from

28:06 transponders uh with with their existing router footprint uh and lastly is

28:11 coherent 100 Gig ZR by eye you know generally when we talk to customers about their access and Edge topologies

28:17 where we expect 100 Gig ZR will have a lot of interest buye Optics are very frequently in the conversation right uh

28:24 for those who are unfamiliar typically your transceivers are duplex meaning there is one fiber strand for transmit

28:30 and another separate fiber strand for receive and the problem is of course that fiber is not always readily

28:36 available not always in good quality or or cheaply accessible so there's an incentive especially in cost sensitive

28:42 access and Edge deployments to optimize fiber utilization using byy Optics which both transmit and receive on a single

28:49 fiber effectively doubling your fiber capacity so this principle is also extending to IP over dwdm deployments

28:56 where customers with scarce fiber may not only wish to optimize fiber capacity

29:01 by using dwdm to MX multiple wavelengths together but also eliminate the need for a fiber pair using by coherent Optics so

29:09 you can see the the top diagram shows a typical dwdm IP over dwdm topology where your mxed signal is going over two

29:16 fibers versus the diagram on the bottom where your mxed signal goes over a single fiber with multiple trans uh

29:22 transmits and receives happening on the same fiber um you know putting all this together right uh kind of that although

29:29 the evolution of 100 Gig ZR offerings will undoubtedly give operators the tools that they need to address the

29:35 specific demands of their use cases it's really important to acknowledge that all of these enh enhancements uh come at the

29:41 expense of increased power consumption and heat dissipation right the these enhancements don't come for free so

29:47 meaning older Brownfield 100 Gig platforms may not necessarily be able to support them uh without restrictions and

29:53 that's again another big reason why we at Juniper are placing so much emphasis on on not only building uh systems with

30:00 the power and thermal margin to service the Optics requirements of today but also the evolving Optics ecosystem down

30:07 the road with that I'll uh think that's my last slide I'll go ahead and pass it to

30:15 Annabelle all right thank you thank you m and with that we're going to uh my

30:23 section which uh is related to best practices like I ideally these new

30:28 technologies always have some nuances that it's important for us to look at so the first thing I want you to keep in

30:36 mind is that the first thing I want to keep

30:41 in mind is new transceivers require powerful tools like always when you are

30:48 thinking on implementing or deploying these new technologies you have to understand that there's a specific

30:55 tools that are needed for it

31:01 so the first thing that I want you to keep in mind is clean and inspect is very important still the second thing is

31:08 validation po points it's important for you to identify what your where you're

31:14 going to validate in your network like you have to identify what are the positions or

31:21 what are the elements that you are going to validate then once you

31:26 identify what are the points which parameters you're going to validate I'm going to talk a little bit about the

31:32 most important parameters amog and Paul already mention them but I want to highlight on how you could do that and

31:39 finally once you have your network deployed or if you're thinking on deployment deploying it what are the

31:45 elements that you can validate in there so with that I want to start with

31:51 or story forever that is cleaning and inspect for coh

31:58 transceivers to keep their characteristics to keep sending information to very long distances to

32:05 use these new uh modulation characteristics that they have for these

32:11 modulations to work as they intend to it's important for you to keep cleaning

32:17 and inspecting your connectors I know we have uh gone through H this burden for a

32:23 long time but in in in terms of coherent technology Sy still need it you have to

32:29 make sure you inspect your connectors to make sure not only if it's dirty but if

32:34 it's not damaged the second point is you have to make sure you inspect

32:40 because you have to make sure you're not connecting an an an a a dirty connector

32:46 but also you can inspect the transceiver itself with these type of technologies that are very complex and they require a

32:54 specific H treatment is very important to even inspect the transceiver and finally it's important

33:02 to inspect also because sometimes when you clean it's not completely clean so

33:08 one of the first best practices that I can give you talking about coherent transceivers still clean and inspect you

33:15 should really see how much time in the interrupts with different organization we take doing this it's it's something

33:23 that is still very important the next uh element that I want you to highlight is you have to

33:30 identify where you are going to validate before you start either deploying or

33:37 troubleshooting the first ER point that I strongly suggest to do is validate

33:44 your transceivers itself like in a loop back like take a test equipment put your

33:50 transceiver and loop it back and do some H generate some traffic because that

33:57 will help you to identify what are the parameters Associated to the vendor that

34:03 you are acquiring or you are working with we know every vendor had a

34:08 different technology or different implementation so it's important that you identify what are the specific

34:13 characteristics of the parameter of the transceivers that you are using I'm going to talk about what is specific to

34:20 measure but make sure you also validate your transceivers look back to

34:25 themselves so the next step is to validate those transceivers interconnected or deployed into a

34:32 network like a mo mentioned there's very different ways you can implement this

34:38 technology so one of the first elements that I ER suggest is to validate your

34:44 transceiver connected to a network but with coherent transceivers there was an

34:50 additional element that we have we didn't have before with a coherent transceiver you can completely directly

34:57 connect the tester to your network so that we couldn't do before so with the

35:02 this type of Technology definitely adding a new uh point of validation

35:08 where is basically connecting your tester to the to the MOX de MOX or amplifier it's very very

35:14 important and the last point is identify if you still once you validated your

35:21 transceiver and then you validated your your coherent network if you still want to see how the traffic goes from your

35:29 ethernet Network to your transport Network so that's still being done

35:34 through a switch and router but it's important for you to also characterize how your traffic is going to pass

35:40 through a the switch and router so with this keep in mind where

35:46 are you validating the next element that I want to highlight is what the first test when

35:53 you take your transceiver and you want to know how this transceiver behaves by itself or in the network there's a

36:01 specific parameters Associated to coherent technology that I I totally

36:06 suggest to validate the first one is application support right now a lot of

36:14 uh Technologies a lot of configurations a lot of parameters are put into the

36:20 transceivers transceivers are now capable of doing a lot of things are accomplishing a lot of stuff so it's

36:26 important for you to buy validate their the application table from these transceivers where you're going to see

36:32 which rates they support which specific implementation that support so having a

36:38 tool that help you to see that is very important to identify what is the particular application that your

36:44 transceiver is going to use another element that is very important when you're validating the

36:51 network or validating this type of transceivers is a monit the power

36:58 consumption as Paul mentioned this type of transceivers can go up to 6.5 Watts

37:03 so monitoring this value is important because you could monitor it inside of a

37:09 lab or you could monitor it in the in the field so it could variate a little bit so that's why it's important for you

37:15 to know what are the the ranges from your transceiver and how it's behaving be because in some cases very specific

37:23 cases we've seen some modifications on that one so so it's important that you

37:30 identify these parameters and see if there's any variation in power power

37:36 consumption and of course the association to it the temperature monitoring as you monitor uh Power the

37:44 temperature has always something to be a monitored next to it because they are

37:49 very Associated so if you see that your temperature is going higher and higher

37:55 that's not something H healthy so so it's something that is important to have

38:00 a look another very important element is the wavelength management like this type

38:05 of transceivers as as uh my colleagues mentioned they can have different wavelengths they can have different GDs

38:12 they can have different channels so it's important for you to identify what are

38:17 the grids that the transceiver that you're using support what are the one that you're going to be using how the

38:23 transer behaves when you change them so having a tool that help you change that

38:28 in a very easy way and in the two scenarios when you are in your lab or when you're in the field is very

38:37 important another element is the optical TX power one of the biggest benefits of

38:44 the coherent transceivers as as Paul mentioned is helping information to travel long distances but this requires

38:51 a lot of power right so in some specific cases and depending on the

38:57 implementation transceivers allow you to boost the the the the power that they

39:03 are generating so having tools that enable enable you

39:09 to a use this feature is very important so it's important to be able

39:16 to test how the transit behaves when you boost the power as well as how it behaves when you send it and how you

39:22 receive it so having tools that help you monitor the power in both directions TX and NX is very

39:29 important and finally as uh we already mentioned coherent transceivers bring a

39:35 lot of benefits and they also have another uh electrical component inside

39:40 that's called DSP that one brings a lot of benefits and one of them is the

39:45 capability of generating measurements right in the transceivers the

39:50 transceivers now are capable of measuring a lot of measuring a lot of parameters and one of them is chromatic

39:57 exeron osnr that help us to know more details about how the link is working so

40:04 having a tool that help you monitor those parameters in real time is very important because this is something we

40:10 could have never done before as as my colleagues mentioned this is bringing

40:16 another ER element of information that provides a better visibility on your

40:22 network so having this information or the capability of monitoring this information is

40:29 important and the final element I want to highlight is once you validated your

40:35 transceiver how you validate your link how you validate a your deployment or

40:41 how you troubleshoot a link that is having problems I would say that I would monitor or what we suggest we monitor

40:48 two very important parameters be error rate and osnr that is signal to noise

40:54 noise radio you can do it with different tools where we allow you to or the tools

41:00 allow you to run traffic over a specific period of time and in that period of

41:05 time you identify if you have errors that's basically the bit error rate test uh that we still suggest that for this

41:12 technology you still do and also you correlate the V rate with another

41:17 parameter that is the signal to noise ratio but here we have uh two variants

41:22 the one that I mentioned you can really read from the transceiver but on top of

41:27 that that is a very good tool I we strongly suggest you to to keep H using

41:33 we also suggest you to do real measurements with an optical Spectrum

41:38 analyzer that will help you to also know what are the optical values and in that way have a sense on how your network is

41:46 behaving between the electrical values and the optical values so with these tools having clean ER cleaning

41:53 inspecting sending traffic validating your parameters and have having uh your

41:59 identification of signal and noise ratio you're going to have a very healthy

42:05 Network and as you saw all of these elements are very interesting and all of these Technologies is very exciting and

42:12 all the team here in this in this webinar am Mo Paul and all the team that is working with us we decided okay this

42:18 is very good technology it's something very interesting but we would like to see it in action we would really like to

42:25 see if all of these we're saying is real so with that we it took the the the or

42:32 and inspired by all the things that that we are experiencing with oif and other

42:40 organizations we said let's do an interrupt test between Juniper precision

42:46 and XO and let's see this working so that's why we created a very nice video

42:52 that we we're going to show you right now where you're going to see how

42:57 um this technology is real it works and how all the benefits we mentioned are

43:03 put into practice so with that uh I think we're going to see the video

43:14 now hello and welcome to this 100 ZR collaborative demonstration by exfo

43:20 Juniper Networks and Precision Optical Technologies we' built a small scale demo to Showcase a real life example for

43:28 testing and deploying 100gr optics for Network operators we've got a couple of

43:33 XO test devices to generate traffic 100 ZR transceivers an eight channel

43:40 mdmx a spool of 75 kilometers of single mode fiber and a juniper host

43:47 platform Central to this system design are footprint optimized 100gr Optics

43:53 that fit within the qsfp28 form factor we're featuring a total of 400 ZR

43:58 plugable Optics two from juniper and two from Precision Optical Technologies our

44:04 test platforms include two of the xfo ftvx 88260 portable units so I have in

44:10 front of me each with a single 100 Gig Port where we have one of the transceivers of juniper and Precision

44:16 Optics plugged in respectively sending traffic by directionally the XO FTB 88260 module is

44:23 the most compact tester on the market it does support gig ZR qsfp28 transceiver

44:29 validation and provides a variety of statistics and values like Optical signal noise ratio uh CD PMD as well as

44:37 the capacity to configure the characteristics of the client and the transmission power these Optics are

44:42 intended for Long Reach applications either in unamplified or Amplified networks with reaches up to 300

44:49 kilometers when paired with an optical line system in our demo here these transceivers are being connected to an

44:57 Channel Max with 100 GHz spacing and we're using it channels 29 and 31 on

45:03 each side being run over 75 kilometers of fiber the MX in another eight channel

45:09 MX that connects to the Juniper ACX router uh hosting the peer side 100 Gig

45:15 ZR Optics the success of Ip over dwdm requires systems designed mindfully to

45:20 support the power and thermal management requirements of pluggable coherent Optics the Juniper ACX 7024 part of

45:27 Juniper's ACX 7000 Cloud Metro portfolio is an access and aggregation router that delivers 24 multi- rate lowp speed ports

45:34 for Downstream connectivity and four 100 Gig ports with uncompromising support for 100 Gig ZR coherent uplinks thanks

45:41 to Juniper's focus on standards the junos operating system leverages cus host to module interface specifications

45:47 to both configure and monitor key Optical parameters on coherent dwdm pluga bles additionally with juniper

45:54 support for open apis operators can extract relevant Optical information from the router and consume them using

46:00 open source visualization tools like raana one of the key advantages of 100 Gig ZR Optics with the ciss is Advanced

46:06 monitoring parameters such as CD and osnr and also the possibility to natively tune the frequency to desired

46:13 channels via the host as we're going to do in the xfo and Juniper hosts as we tune to the itu channels 33 and 35 for

46:22 host platforms that do not support simis yet or have Legacy software the one be upgraded we're also going to show the

46:29 tuning capability of both Optics uh for sfff manag optics for these 100 ZR

46:35 transceivers external tuning will need to be performed using a host like the Precision tn100 qsd management

46:42 module after tuning both transceivers to a different Channel we will change the physical MX connections to channel 33

46:48 and 35 on both sides causing the links to go up again and traffic to start flowing bidirectionally on both xfo and

46:54 Juniper host the availability of qsfp28 100 Gig ZR Solutions Builds on the

46:59 success of 400 gig zrzr plus extending the compelling TCO benefits of Ip over dwdm to new Metro Data Center and

47:07 Enterprise use cases with real world alignment of low power and small form factor coherent pluga bles purpose built

47:13 routers and test systems and robust industry standardization the time is now for deploying a converged Optical

47:18 routing architecture

47:25 excellent well I I think with that we we were able to see the video and and and

47:30 as you saw we were very inspired by all the collaboration from the industry and it was super cool to see how we were

47:38 able to bring together this video with for you to show you how this technology

47:43 actually works and in that same line I would like to invite you to come and see all of us

47:51 all of these very good uh ER elements that we mention from from ER the company

47:59 Precision from juniper and from xfo I totally invite you to come and see all

48:04 of us to ofc a 2025 that is going to happen from

48:09 April April 1st to 3rd all of us we have different uh boots as well as we some

48:16 some of us we are participating on the oif and internet Alliance uh interconnection so please come and visit

48:23 us please come and see this technology that is very important for the industry and don't miss it out definitely we're

48:29 going to be very happy to see you there and I think with that we conclude

48:34 the presentation and we can go to the questions I give the microphone to Sean

48:40 great yeah thank you no that was a great uh that's kind of a neat thing to put

48:45 into a webinar I haven't seen someone do that in a while but again you guys are talking about real world issues and uh

48:52 you know again the the collaboration between some really great uh uh players here okay great um I guess I'll jump

48:59 into some questions here um let's see um I'll just kind of randomly take

49:05 some from the audience here um all right here we go so we have seen the same promises with 400g ZR Plus for almost

49:12 five years now and the real deployments are very limited until now what could be the differentiator in 100g ZR that makes

49:18 it a real viable Deployable solution any of you want to take that first

49:25 yeah yeah I I got I can get started I I think the the form factor here the qsp

49:30 28 is is absolutely key um I think also some of the learnings that happen with

49:36 400 G ZR ZR Plus have been applied here like um initially u a lot of the 400 ZR

49:44 was meant for data center inter connect uh the cloud hyperscalers and for example when the

49:50 service providers start looking into that solution they said well we need something more than that we need and that's how zr+ how the differ itations

49:57 for TX output power that we've mentioned up to Z dbm two even plus 4 dbm came

50:03 into play so for example 100 Gig ZR you can see from the get-go we already taken

50:09 that into account we have this 100 grr with low TX High TX power we have these options for hardened temperature as Mo

50:16 mentioned CM so I think uh the learnings that happen on on on those ZZR

50:22 developments are being appli on 100 Gig ZR and I think that's what going to make this

50:32 success I don't hear you sir but maybe I can take it I I wanted

50:38 to add also to that I think one of the elements is 100 set is a power

50:44 consumption it's really lower compared to to what 400 gig had right so reducing

50:51 that amount of power consumption is going to allow the deployments to be done farther and farther from the core

50:57 and from the Metro applications so I think that will help to age applications and some other that that are going to

51:04 come in the future right great great yeah I had myself muted thank you um

51:10 great okay I guess we'll just keep rolling along for another 10 minutes unless someone else wants to add but uh

51:17 yeah just the audience keep rolling along the questions we get some great stuff here um all right let's see uh let

51:25 see uh one second um let's see is this a question

51:31 for Juniper does Juniper plan to support qsfp 2800 gzr in ACX 7100 48 l in AC

51:43 PTX uh 10,136 MC models if yes when I think

51:48 someone's asking you guys a specific product question but I don't know if you want to address that yeah um so good

51:55 news 100 gigs is already supported on both of those platforms um in the the release we just put out at the end of

52:01 last year 24.4 R1 um just a a caveat to note uh the ACX routers um come in two

52:09 two flavors one of them is uh front to back air flow the other one is back to front airflow for uh based on you know

52:16 the the air flow requirements of different data center deployments um so 100 Gig Z is supported on the front to

52:22 back airf flow but the back to front introduces preheat basically that comes to the the ports right so um I think

52:30 you'll find that any back to- front airflow system in the entire industry won't support coherent Optics because of

52:37 that preheat contribution that goes to the front panel um but short answer yes we do support

52:42 it great great that's good to know and I guess everybody could probably see that uh ofc um here here's one question that

52:50 you know I typically ask all of you guys when I'm doing an interview or as I'm trying to learn about this industry more

52:55 you know you know why is IP over dwdm getting Traction in the market now you know what's different than the past I

53:01 think a couple of you did address address the initial uh you know applications for for this this

53:06 technology but it seems like you know more and more we see more carriers you know deploy it really and several

53:13 several the customers that all of all three of you serve I know are looking at this you mentioned Alisa I think too but

53:20 yeah I don't know who wants to jump on that or all three of you want to weigh in yeah yeah I mean maybe I can start um so

53:29 I think that the it's a good point because IP over dwdm Is Not A New Concept right this this is something that's been debated in the industry for

53:35 a long time but it's only kind of in the last few years that we're seeing actual adoption so I think there's probably

53:41 three reasons that come to mind um first is standardization I think we had plugable you know cfp2 dcos in the

53:48 market um these were proprietary interfaces right meaning that the overall solution is not open and

53:53 interoperable and you're kind of locked in you're forced to lock into a small proprietary ecosystem of vendors which

53:59 translates to high prices and a lack lack of deployment flexibility um and second I think as we all kind of touched

54:05 on in our presentations is the footprint right uh although we had stuff like the cfp2 dco um it was physically larger and

54:13 consumed a ton of power which means you required you not only required special line cards on your router to support

54:20 cfp2 form factors but you also end up sacrifi sacrificing you know your your density on on a line card or router or

54:27 switch um because of having a a larger physical footprint um and I think the

54:33 last thing I'll add is is is as I mentioned system design right routers in the past were never really designed or

54:39 were were thinking about the power and heat requirements of high performance plug bles right it was always we use our

54:45 client Optics we're going to hand off to a a dedicated transponder uh and and that's it right we're never going to

54:51 support these you know 20 25 watt now 30 Watts at 800 gig um this is all to us

54:56 right so it's only now that we're starting to see system vendors actually being being mindful in the design of

55:02 systems to to handle the requirements of pluga bles like 100 400 or even 800 ZR

55:08 ZR plus okay great uh XO or you know Delon

55:14 want to take a shot as well or yeah well I think for me on my side

55:21 what I've seen from from my interaction with customers is one of the biggest benefits is the longer distance right

55:26 that we can actually get a very long distances links as well as for example

55:35 in testing now you could actually connect the tester directly to the transport Network so in in in my

55:42 perspective of testing I think those would be the two very important elements that that are seeing are I'm seeing the

55:50 traction of this new technology great okay um all right

55:55 wonderful all right for no one else I'll just keep moving along with some questions here um

56:01 here's an interesting one that's been proposed what are the most common issues encountered when deploying coherent transceivers I mean you know obviously

56:08 it's it's it's a technology that's been around for quite a while but now it seems like we're getting more momentum and you know what what are your you know

56:15 are you guys seeing out in the field when you're working with customers deploying these things and yeah that may be a good one to put out

56:22 yeah well something that I don't know Expo has been advocate for a long time about the clean and inspection I think

56:29 everybody has heard our message about it but I have to say that we've been very

56:35 very um grateful to participate in a lot of interrupts when this technology was

56:42 being Define uh since 800 400 100 we've been involved in a lot of that and it's

56:49 incredible how the the biggest difference you get when you clean and inspect your connector and when you

56:55 don't so I would say for me the lesson number one for this one is that and also

57:03 make sure you identify what's the temperature and power consumption of your transer that is

57:09 something at the beginning of the coherent transer era that was the main

57:15 thing that we were asking people to do and I think is still important to

57:20 monitor how your transer is going to behave either in the switch either by itself to ciz that those

57:29 parameters great yeah on yeah on our side obviously a position we're we're

57:34 dealing with system integration every day right so uh truly with coherent transceivers is just a next step like a

57:42 typically IMDd nerz Optics you you could like plug it in a hose do a physical

57:48 loop back if it works it works right um with with coherent Optics that there's that next level right like the what

57:54 Anabel just mentioned the consumption and move mention on on the heat dissipation so we encourage all all

58:01 customers we we are testing this to to either the a lab trial or a field trial

58:07 do extended testing for these Optics to make sure uh these are not just like a

58:12 oneoff case and and specific take into account to not like fully load some of

58:18 these switches or host with several of these like take it one at a time put it

58:23 one put it two make sure uh the platform especially for some of the Legacy host

58:29 uh we got to be careful with how many of these ones we load uh before we do full scale Mass

58:36 deployments okay great um another qu there's been several questions coming in

58:41 about uh B bidi coherent 100g ZR ZR plus

58:47 uh platforms I guess you know are they available in shipping now and you know kind of you know what's everyone's

58:53 perspective on that on that topic I know it seems like some folks really want to know a lot more about that

58:59 yeah yeah so what one thing that's interesting in 100 Gig and by Optics

59:05 historically in IMDd where just like one single LC connector uh transmitting at

59:10 different frequencies in ceden Optics uh that's typically not possible unless the

59:16 Cohen transceiver has a dual laser manufactured as part of the packaging

59:22 the reason is that the local oscillator using the receiver for the transmitted

59:27 signal is also used as a reference point for the receiving signal when you have only a single laser so truly to do 100

59:34 Gig by and more than 100 Gig by is 100 Gig single fiber dwdm deployment um that

59:41 requires this transceivers to have a dual laser configuration that's a product that is in the road map but is

59:47 not available yet so it's that something that we expect to come at some point is not a product that's available yet okay

59:56 all right we're kind of running up up against the hour but uh and we there's a lot of questions here we couldn't get to

1:00:01 but we'll have everyone we'll have the uh speakers follow up with everyone but if I could just quickly for every speaker here could I just have each of

1:00:07 you just give a couple quick final thoughts and then we'll close out the uh close out the uh webinar uh if anyone

1:00:13 wants to start first real quick yeah I can have a quick start oh

1:00:20 go ahead sorry um yeah no I think it's a it's a it's a very exciting time I think

1:00:27 in this particular domain I think we're finally starting to see uh all the key building blocks come together um I think

1:00:33 the 400 Gig Generation gave us a lot of learnings both from how to design systems uh what the actual operator

1:00:40 requirements are to actually deploy this technology now you know as the technology evolves to 100 and 800 um all

1:00:47 the building blocks are coming together right the pluga bles the Open Line systems the the routers and switches uh

1:00:53 and and the architecture is real right I I I know that one of the questions was about like uh you know we don't see a

1:00:59 lot of deployments I I disagree right I I think there are actually quite a bit of deployments uh just the complexity of

1:01:05 Any Given topology may take operators some time to actually migrate architectures uh but it's definitely

1:01:11 being adopted or at least tested uh in large scale you know cloud and service provider Network so very exciting time

1:01:18 and it's great to see that that finally the ecosystem is is is ready to to make this happen for

1:01:24 customers great and and a yeah yeah I I agree with a mug I think

1:01:30 is exciting I think this product like we can hear from all the customers every everyone uh wants to get the hands on

1:01:37 these Optics everyone wants to be testing and I think the the takeaway is this like test test test like get these

1:01:44 parts out in the lab get the field integrate into the system make sure it

1:01:49 fits the solution I'm sure that makes Annaville happy for sure exactly

1:01:56 I love that message I love that message and if you want if I can add my side too

1:02:01 like I totally suggest that test test and also keep in mind that it's not just

1:02:07 one solution you can have your inspection you can have your traffic generator you can have your Osa like

1:02:13 it's an integral testing like you have to test in several areas and also for this particular solution consider

1:02:20 portable solution the the most compact the better that will help you to move

1:02:26 around all the needs you have right so basically I think the technology is

1:02:31 ready the technology is real as you could see it in the video we were very excited to do that and you can see it

1:02:38 real working also at ofc great well with that I me I think

1:02:44 that's a great way to end this and uh thanks all three speakers uh some really great uh insights today you know test

1:02:50 and validate and you know I think it's going to be interesting to see kind of the Innovation that from these three companies and others um at ofc uh you

1:02:58 know late next month and uh look forward to seeing it myself thank you very much thanks to the audience and uh I hope

1:03:04 everybody has a great day thank you very much thank you for you ATC I'll be there for to to meet you guys all right sounds

1:03:12 good see you soon bye- bye hello and welcome to this 100 ZR

1:03:19 collaborative demonstration by exfo Juniper Networks and Precision Optical

1:03:24 Technologies we've bu built a small scale demo to Showcase a real life example for testing and deploying 100g

1:03:31 optics for Network operators we've got a couple of XO test devices to generate

1:03:38 traffic 100 ZR transceivers an 8 channel maxd Max spool of 75 kilom of single

1:03:46 mode fiber and a juniper host platform Central to this system design

1:03:52 are footprint optimized 100gr Optics that feed within the qsa 28 form factor

1:03:58 we're featuring a total of 400 ZR plugable Optics two from juniper and two

1:04:03 from Precision Optical Technologies our test platforms include two of the xfo

1:04:08 ftvx 88260 portable units who I have in front of me each with a single 100 Gig

1:04:15 Port where we have one of the transceivers of juniper and Precision Optics plugged in respectively sending

1:04:21 traffic by directionally the Expo FTB 88260 module is the most compact tester on the market

1:04:28 it does support 100 Gig ZR qsfp28 transceiver validation and provides a

1:04:33 variety of statistics and values like Optical signal noise ratio uh CD PMD as

1:04:39 well as the capacity to configure the characteristics of the client and the transmission power these Optics are

1:04:44 intended for Long Reach applications either in unamplified or Amplified networks with reaches up to 300 kmet

1:04:52 when paired with an optical line system in our for demo here these transceivers are being connected to an8 channel Max

1:05:00 with 100 GHz spacing and we're using itu channels 29 and 31 on each site being

1:05:06 run over 75 kmers of fiber the Mar in another eight channel Marx that connects

1:05:12 to the Juniper ACX router uh hosting the pier side 100 Gig ZR Optics the success

1:05:19 of Ip over dwdm requires systems designed mindfully to support the power and thermal management requirements of

1:05:25 plug able coherent Optics the Juniper ACX 724 part of Juniper's ACX 7000 Cloud

1:05:31 Metro portfolio is an access and aggregation router that delivers 24 multi- rate lowp speed ports for

1:05:37 Downstream connectivity and four 100 Gig ports with uncompromising support for 100 Gig ZR coherent uplinks thanks to

1:05:44 Juniper's focus on standards the junos operating system leverages seus host to module interface specifications to both

1:05:50 configure and monitor key Optical parameters on coherent DWD imp pluga bles additionally with Juniper support

1:05:56 for open apis operators can extract relevant Optical information from the router and consume them using open

1:06:02 source visualization tools like raana one of the key advantages of 100 Gig ZR Optics with the ciss is Advanced

1:06:08 monitoring parameters such as CD and osnr and also the possibility to natively tune the frequency to desired

1:06:15 channels via the host as we're going to do in the xfo and Juniper host as we tune to the itu channels 33 and 35 for

1:06:24 host platforms that do not support simis yet or have Legacy software that won't be upgraded we're also going to show the

1:06:31 tuning capability of both Optics uh for sff manage optics for these 100 ZR

1:06:37 transceivers external tuning will need to be performed using a host like the Precision tn100 qsd management

1:06:44 module after tuning both transceivers to a different Channel we will change the physical MX connections to channel 33

1:06:50 and 35 on both sides causing the links to go up again and traffic to start flowing bidirectionally on both xfo and

1:06:56 Juniper host the availability of qsfp28 100 Gig ZR Solutions Builds on the

1:07:01 success of 400 gig zrzr plus extending the compelling TCO benefits of Ip over dwdm to new Metro Data Center and

1:07:09 Enterprise use cases with real world alignment of low power and small form factor coherent pluga bles purpose-built

1:07:15 routers and test systems and robust industry standardization the time is now for deploying a converged Optical

1:07:20 routing architecture

1:07:27 excellent well I think with that we we were able to see the video and and and

1:07:32 as you saw we were very inspired by all the collaboration from the industry and it was super cool to see how we were

1:07:40 able to bring together this video with for you to show you how this technology

1:07:45 actually works and in that same line I would like to invite you to come and see all of us

1:07:53 all of these very good uh ER elements that we mention from ER the

1:08:00 company Precision from juniper and from xfo I totally invite you to come and see

1:08:06 all of us to ofc ER 2025 that is going to happen from

1:08:11 April April 1st to 3rd all of us we have different uh boots as well as we some

1:08:18 some of us we are participating on the oif and ethernet Alliance uh interconnection so please come and visit

1:08:24 us please come and see this technology that is very important for the industry and don't miss it out definitely we're

1:08:31 going to be very happy to see you there and I think with that we conclude

1:08:36 the presentation and we can go to the questions I give the microphone to Shan

1:08:42 great yeah thank you no that was a great uh that's kind of a neat thing to put

1:08:47 into a webinar haven't seen someone do that in a while but again you guys are talking about real world issues and uh

1:08:54 you know again the the coll collaboration between some really great uh uh players here okay great um I guess

1:09:00 I'll jump into some questions here um let's see um I'll just kind of randomly take

1:09:07 some from the audience here um all right here we go so we have seen the same promises with 400g ZR Plus for almost 5

1:09:14 years now and the real deployments are very limited until now what could be the differentiator in 100g ZR that makes it

1:09:21 a real viable Deployable solution and you want to take that first

1:09:27 yeah yeah I can I can get started I I think the the form factor here the qsp

1:09:32 28 is is absolutely key um I think also some of the learnings that happen with

1:09:38 400 G ZR ZR Plus have been applied here like um initially um a lot of the 400 ZR

1:09:46 was meant for data center inter connect uh the cloud hyperscalers and for example when the

1:09:52 service providers start looking into that solution they said well we need something more than that we need and that's how zr+ how the different

1:09:58 interations for TX output power that we've mentioned up to Z dbm to even plus

1:10:04 4 dbm came into play so for example 100 Gig Z you can see from the GetGo we

1:10:10 already taken that into can we have this 100gr with low TX High TX power we have

1:10:15 these options for hardened temperature as Mo mentioned CM so I think uh the

1:10:21 learnings that happen on on on those ZZR developments are being applied 100 big

1:10:26 and I think that's what going to make this a huge

1:10:34 success I don't hear you son but maybe I can take it I I wanted

1:10:40 to add also to that I think one of the elements is 100 setr is a power

1:10:46 consumption it's really lower compared to to what 400 gig had right so reducing

1:10:53 that amount of power consumption is going to allow the deployments to be done farther and farther from the core

1:10:59 and from the Metro application so I think that will help to H applications and some other that that are going to

1:11:06 come in the future right great great yeah I have myself muted thank you um

1:11:12 great okay I guess we'll just keep rolling along for another 10 minutes unless someone else wants to add but uh

1:11:19 yeah just the audience keep rolling along the questions we get some great stuff here um all right let's see

1:11:26 uh let's see uh one second um let's see is this a question

1:11:33 for Juniper does Juniper planed support qsfp 2800 gzr and ACX 7100 48 L and AC

1:11:45 PTX uh 10,136 M AC models if yes when I think

1:11:50 someone's asking you guys a specific product question but I don't know if you want to address that so yeah um so good

1:11:57 news 100 Gig ZR is already supported on both of those platforms um in the the release we just put out at the end of

1:12:03 last year 24.4 R1 um just a a caveat to note uh the ACX routers um come in two

1:12:11 two flavors one of them is uh front to back air flow the other one is back to front air flow for uh based on you know

1:12:18 the the air flow requirements of different data center deployments um so 100 gigs Z is supported on the front to

1:12:24 back are flow but the back to front introduces preheat basically that comes to the the ports right so um I think

1:12:32 you'll find that any back to- front airflow system in the entire industry won't support coherent Optics because of

1:12:38 that preheat contribution that goes to the front panel um but short answer yes we do support

1:12:44 it great great that's good to know and I guess everybody could probably see that uh osc um here here's one question that

1:12:51 you know I typically ask all of you guys when I'm doing an interview or as I'm trying to learn about this industry more

1:12:57 you know you know why is IP over dwdm getting Traction in the market now you know what's different than the past I

1:13:03 think a couple of you did address address the initial uh you know applications for for this this

1:13:08 technology but it seems like you know more and more we see more carriers you know deploy it really and SE several

1:13:15 several of the customers that all of all three of you serve I know are looking at this you mentioned Alisa I think too but

1:13:22 yeah I don't know who wants to jump on that or all three of you want to weigh in yeah yeah I mean maybe I can start um so

1:13:31 I think that the it's a good point because IP over dwdm Is Not A New Concept right this this is something that's been debated in the industry for

1:13:37 a long time but it's only kind of in the last few years that we're seeing actual adoption so I think there's probably

1:13:43 three reasons that come to mind um first is standardization I think we had pluggable you know cfp2 dcos on the

1:13:50 market um these were proprietary interfaces right meaning that the overall solution is not open and

1:13:55 interoperable and you're kind of locked in you're forced to lock into a small proprietary ecosystem of vendors which

1:14:01 translates to high prices and a lack lack of deployment flexibility um and second I think as we all kind of touched

1:14:07 on in our presentations is the footprint right uh although we had stuff like the cfp2 dco um it was physically larger and

1:14:15 consumed a ton of power which means you required you not only required special line cards on your router to support

1:14:22 cfp2 form factors but you also end up AC sacrificing you know your your density

1:14:27 on on a line card or router or switch um because of having a a larger physical

1:14:33 footprint um and I think the last thing I'll add is is is as I mentioned system design right routers in the past were

1:14:39 never really designed or we we thinking about the power and heat requirements of high performance plug bles right it was

1:14:46 always we use our client Optics we're going to handoff to a a dedicated transponder uh and and that's it right

1:14:52 we're never going to support these you know 20 2 5 watt now 30 Watts at 800 gig

1:14:57 um this is all new to us right so it's only now that we're starting to see system vendors actually being being

1:15:02 mindful in the design of systems to to handle the requirements of pluga bles like 100 400 or even 800 ZR ZR

1:15:11 plus okay great uh XO or you know building want to take a shot as well

1:15:20 or yeah well I think for me on my side what I've seen from from my interaction

1:15:25 with customers one of the biggest benefits is the longer distances right that we can actually get a very long

1:15:33 distances links as well as for example in testing now you could actually

1:15:38 connect the tester directly to the transport Network so in in in my

1:15:44 perspective of testing I think those would be the two very important elements that that are seeing are I'm seeing the

1:15:52 traction of this new technology great okay um all right wonderful all

1:15:58 right if any one else I'll just keep moving along with some questions here um here's an interesting one that's been

1:16:04 proposed what are the most common issues encountered when deploying coherent transceivers I mean you know obviously

1:16:10 it's it's it's a technology that's been around for quite a while but now it seems like we're getting more momentum and you know what what are your you know

1:16:17 are you guys seeing out in the field when you're working with customers deploying these things and yeah that might be a good one to put out yeah

1:16:25 well something that I don't know EXO has been advocate for a long time about the clean and inspection I think everybody

1:16:31 has heard our message about it but I have to say that we've been very very um

1:16:40 grateful to participating a lot of interrupts when this technology was being defined uh since 800 400 100 we've been

1:16:48 in involved in a lot of that and it's incredible how the the biggest difference you get when you clean and

1:16:55 inspect your connector and when you don't so I would say for me the lesson

1:17:01 number one for this one is that and also make sure you identify what's the

1:17:07 temperature and power consumption of your transceiver that is something at the beginning of the

1:17:14 coherent transceiver era that was the main thing that we were asking people to

1:17:20 do and I think is still important to monitor how your transer is going to behave either in the switch either by

1:17:26 itself to characterize that those parameters great yeah on yeah on our

1:17:35 side obviously OPP position we're we're dealing with system integration every day right so uh truly with with coherent

1:17:41 transceivers is just a next step like a typically IMDd nrz Optics you you could

1:17:47 like plug it in a hose do a physical Lo back if it works it works right um with

1:17:53 with coherent Optics that there's that Next Level right like the what Annabel just mentioned the power consumption and

1:17:59 move mention on on the heat dissipation so we encourage all all customers we

1:18:04 we're testing this to to either the a lab trial or a field trial do extended

1:18:10 testing for these Optics to make sure uh these are not just like a oneoff case

1:18:15 and and specific take into account to not like fully load some of these

1:18:20 switches or host with several of these like take it one time put it one put it

1:18:26 two make sure uh the platform especially for some of the Legacy host uh we got to

1:18:31 be careful with how many of these ones we load uh before we do full scale Mass

1:18:38 deployments okay great um another qu there's been several questions coming in

1:18:43 about uh B bidi coherent 100g ZR ZR plus

1:18:49 uh platforms I guess you know are they available in shipping now and you know kind of you know what's everyone's

1:18:55 perspective on that on that topic I know it seems like some folks really want to know a lot more about that

1:19:01 yeah yeah so what one thing that's interesting in 100 Gig and by Optics

1:19:07 historically in IMDd were just like one single LC connector uh transmitting at

1:19:12 different frequencies in coherent Optics uh that's typically not possible unless

1:19:18 the Cohan transceiver has a dual laser manufactured as part of the packaging

1:19:24 there is reason is that the local oscillator using the transceiver for the transmitted signal is also used as a

1:19:31 reference point for the receiving signal when you have only a single laser so truly to do 100 Gig by and more than 100

1:19:37 Gig by is 100 Gig single fiber DDM deployment um that requires these

1:19:43 transceivers to have a dual laser configuration that a product that is in the road map but is not available yet so

1:19:50 it's that something that we expect to come at some point not a product that's available yet okay all right we're kind

1:19:58 of running up up against the hour but uh and we there's a lot of questions here that we couldn't get to but we'll have

1:20:03 everyone we'll have the uh speakers follow up with everyone but if I could just quickly for every speaker here could I just have each of you just give

1:20:09 a couple quick final thoughts and then we'll close out the uh close out the uh webinar uh if anyone wants to start

1:20:15 first real quick yeah yeah I can give a quick start oh go

1:20:22 ahead I'm sorry um yeah no I think it's a it's a it's a very exciting time I

1:20:28 think in this particular domain I think we're finally starting to see all the key building blocks come together um I

1:20:35 think the 400 Gig Generation gave us a lot of learnings both from how to design systems what the actual operator

1:20:42 requirements are to actually deploy this technology now you know as the technology evolves to 100 and 800 um all

1:20:49 the building blocks are coming together right the pluga bles the Open Line systems the S the routers and switches

1:20:54 uh and and the architecture is real right I I I know that one of the questions was about like uh you know we

1:21:00 don't see a lot of deployments I I disagree right I I I think there are actually quite a bit of deployments uh

1:21:06 just the complexity of Any Given topology may take operators some time to actually migrate architectures uh but

1:21:12 it's definitely being adopted or at least tested uh in large scale you know cloud and service provider Network so

1:21:19 very exciting time and it's great to see that that finally the ecosystem is is is ready to to make this happen for

1:21:26 customers great and Annabelle and amog yeah yeah I I agree with Aug I think is

1:21:32 exciting I think this product like we can hear from all the customers every everyone uh wants to get the hands on

1:21:39 these Optics everyone wants to be testing and I think the the takeaway is this like test test test like get these

1:21:46 parts out in the lab get the field integrate into the system make sure it

1:21:51 feits the solution I'm sure that makes annab happy sure

1:21:57 exactly I love that message I love that message and if you want if I canot my

1:22:02 side too like I totally suggest that test test and also keep in mind that

1:22:09 it's not just one solution you can have your inspection you can have your traffic generator you can have your Osa

1:22:15 like it's an integral testing like you have to test in several areas and also for this particular solution consider

1:22:22 portable solution the the most compact the better that will help you to move

1:22:27 around all the needs you have right so basically I think the technology is

1:22:33 ready the technology is real as you could see it in the video we were very excited to do that and you can see it

1:22:40 real working also at ofc great well with that I me I think

1:22:46 that's a great way to end this and uh thanks all three speakers uh some really great uh insights today you know test

1:22:52 and validate and you know I think it's going to be interesting to see kind of the Innovation that uh from these three companies and others um at ofc uh you

1:23:00 know late next month and uh look forward to seeing it myself thank you very much thanks to the audience and uh I hope

1:23:06 everybody has a great day thank you very much thank you for you ATC I'll be there for to to meet you

1:23:12 guys all right none's good see you soon bye bye