Raymond James Financial, Inc. (CSCO) Earnings Call Transcript & Summary

Ladies and gentlemen, thank you all for standing by and welcome to the Virtual Gathering Series featuring Cisco Conference call. [Operator Instructions] Please be advised that today's conference call is being recorded. [Operator Instructions] I would now like to hand the conference over to your speaker today, Mr. Simon Leopold. Thank you. Please go ahead, sir.

Great. Thank you very much, folks. Sorry for the delay, but it sounds like the system is under stress with high demand, many folks interested in joining. This is Simon Leopold, Raymond James data infrastructure analyst with the first in our series of virtual gathering. Thanks for joining us today with Cisco's SVP and GM for Optical Systems and Optical Group, Bill Gartner. We also have Emily Hunt from the Investor Relations team on as well. In light of the recent challenges that have limited all of our travel, we hope to provide an opportunity to hear about markets and technology advances because we missed the opportunities to catch up at Mobile World Congress and the OFC show, we hope to gain the kind of insight that would complement what we typically learn on earnings call. We have prepared some questions for Bill and will also offer an opportunity for the folks dialed in to queue for questions at the end. Let me just give you a little bit of background on Bill Gartner. He has the profit and loss responsibility for Cisco's Optical business and is responsible for strategic direction, product development, product positioning and new product introduction. Prior to joining Cisco, Bill had senior management positions at several private optical companies. Before joining Photuris, one of the first to introduce ROADM, Bill was Vice President and General Manager of Lucent's Optical Networking business unit. Let me see -- hand it off to Emily for a couple of disclaimers, and then Bill will offer some opening remarks, and we'll dive into the Q&A. Emily, the floor is yours.

Thanks, Simon. I'd like to remind the audience that today's call relates to Cisco's Optical business only. No new financial information regarding Cisco's overall performance is intended or implied. We may make forward-looking statements, which are subject to risks and uncertainties outlined in detail in our documents filed with the SEC, including our most recent filings of Form 10-K and 10-Q, and actual results may differ from statements made today. With that, I'd now like to hand over to Bill.

Thanks, Emily, and thank you, Simon, for hosting this event. And I'm sorry, we didn't get a chance to meet up face-to-face at OFC. Obviously, the coronavirus has been creating a lot of chaos in our worlds and Cisco is doing everything it possibly can to keep our employees safe and have instituted work from home policies for many of our employees, where it makes sense. And obviously, our supply chain teams are very busy in making sure we can mitigate any supply chain issues as well. Let me just start with a few comments about to frame the -- where we are in the business and to, hopefully, tee up some further discussion and questions from you. I think if you hear Chuck talk or you go to any industry presentation from Cisco, you will hear a lot about software, recurring revenue model, subscription models and businesses like our security business and functions like automation, AppD, things like that have really captured a lot of the conversation for Cisco. But I think one of the things that we can't lose sight of is the fact that our core networking business, which includes routers and switches, optical systems, cable systems, still forms the foundation for many of those software applications. And in fact, there are 3 technology pillars that have been identified by Cisco as sort of the key technologies that we need to be investing in. One of those is software, obviously, but also silicon and optics. And I'm very glad that optics has sort of risen to the level of being 1 of the 3 key pillars that Cisco focuses its investments on. I want to just also frame the world that I play in, in Cisco, which is I own the Optics business and the Optical Systems business. And I want to just separate those in your mind. So as we're having conversations, we're using the right semantics. The Optics business, you can think of as the business that takes place within the data center or within a central office in a service provider environment and it's characterized by 2 attributes. One is that fiber is plentiful, meaning if you add a new switch or you add a new router, you pull new fiber and every port on a router or switch gets its own fiber, and the other attribute is the distances are short, typically less than 10 kilometers. And for that application, we can use transceivers, pluggable transceivers that use direct detect technology, which basically means if the lights on, it's a 1. If the lights off, it's a 0. And the game is, make the light go on and off as fast as you possibly can. And so that's the optics world that we play in. We have a very substantial optics business that serves all of our customer segments, from service provider, web to enterprise and campus applications. On the other side is the Optical Systems business, which you can think of as what happens outside the data center or outside of central office, where we now need to move signals across the city or across a country or even across the ocean. And in that world, fiber is no longer plentiful, now fiber is scarce because I can't ask a service provider to pull a new fiber between L.A. and New York because they happen to add a new router. That fiber is in the ground, and we have to basically leverage that existing investment. And we do that with technologies like DWDM that allow us to put many signals on 1 fiber. So fiber being scarce is one attribute. And the other attribute is if the distances are long, typically 80 kilometers up to several thousand kilometers. So that makes it much more challenging to send the light over long distances. And we have to rely on very sophisticated technologies that are called coherent technologies. And it's much more sophisticated technology used outside the data center versus inside the data center. Now I want to just summarize some of the investments that Cisco has made, and we'll be happy to take some questions on these. But going back to 2010, we invested in a company called CoreOptics, which was our first coherent introduction of technology, and that has served us very well as we've introduced 100 gig, 200 gig, 250 gig with the coherent solutions from CoreOptics. In 2012, we acquired our first silicon photonics company, which was Lightwire, and that also has served us very well. It actually enabled the industry's first 1 terabit line card on a router. And we've continued to invest in that technology. And more recently, in early 2019, we acquired a company called Luxtera. And Luxtera brings us sort of a next-generation of silicon photonics, and they had a very highly automated manufacturing platform that really allows us to manufacture pluggables or optics modules at a very high scale. And then finally, we announced midyear last year, our intention to acquire Acacia, which again is a coherent solution that was targeted outside the data center primarily. And Acacia brings not only coherent technology, but also silicon photonics, and they've put that together in a very creative way that allows us to think about moving some technologies from chassis-based solutions in the optical world into pluggables that would go into a router, and that's a very interesting dynamic for us, and we'll be happy to talk a little bit more about that. Now you may wonder like why does Cisco care about optics? And let me just summarize it with 2 reasons. One is that the cost, the relative cost of optics versus the port is increasing as the bit rate increases. From 10 gig to 100 gig to 400 gig, optics becomes a bigger portion of that spend. So that's one important reason if we're going to remain relevant for our customers and we're going to continue to be able to serve our customers with fully integrated solutions, we have to be able to control some of that cost. And then the second issue, which is a little bit more subtle, is that today, inside the data center, we use these pluggable transceivers that rely on pretty, I'm going to call it simple, although it's not all that simple technology, which is direct detect. Lights on, it's a 1. Lights off, it's a 0. As the bit rate increases, from 100 gig to 400 gig to 800 gig and beyond, sending that signal over a longer distance becomes more challenging. As the bit rate increases in general, physics conspires against us and the distance gets shorter. That's just physics working against us. And so what we expect is that at some point, as the bit rate increases, the ability to send that signal over 2 kilometers or 10 kilometers or 40 kilometers becomes challenging for direct detect, and we'll actually have to use the much more sophisticated coherent technology, maybe even inside 10-kilometer applications or maybe even 2-kilometer applications. And so that motivated our thinking around Acacia to say, hey, this technology may have application more than just outside the data center, it may actually come inside the data center or central office at some point. So that's kind of our high level thesis around these investments. I would say we're continuing to make big investments organically and inorganically in these areas of optics and silicon. We're going to see optics drive architectural transitions inside the data center and outside the data center, that's kind of upside down from where we've always been where optics has kind of followed along. Optics will, in fact, drive some of these transitions. And I don't think there's any company on earth that's making the investment in silicon, optics and software at the level of Cisco. So some of these investments are long-term investments, we won't see benefits of for 3 to 5 years. But I think we're positioning ourselves well for the future of the Internet here. So with that, Simon, hopefully, that would help to at least frame some of the issue and maybe give you some fodder for questions. I'm happy to take any questions you might have.

Yes, sure. So I appreciate that. And certainly, a lot to digest there. Maybe just at the high level, optics seems to be a pretty broad topic for Cisco, inside the data center, longer reach, development in silicon photonics, developments of coherent. I guess, what I'd like to get a sense of is among kind of all these aspects you've highlighted, what are you particularly most excited about in terms of technological breakthroughs and developments?

So I think that -- let me answer it with 2 different things. One is I think silicon photonics is continuing to show that it's a technology that can expand its application. And so we've seen silicon photonics initially was sort of very targeted at short-reach applications. We're now seeing silicon photonics in coherent long-reach applications. So I think there's a lot of very exciting things happening in silicon photonics that expand the domain for that technology, and we're continuing to invest pretty heavily in that area. And then I think the big shift that I think we're going to see is today, we deploy DWDM systems for a long-haul network, like a Verizon network, for instance, or an AT&T or a DT network. And these are chassis that get deployed and the chassis are line cards, and the line cards have things like ROADMs on it, and they also have transponders or coherent on it to deliver the signals into the fiber. Some of those technologies, those coherent technologies are now thanks to advances in ASICs, like the SPEs and like silicon photonics, some of those technologies are going to move from being on a line card to being on a pluggable that can go directly into the router. The 400-gig ZR is the best current example. That's an architectural shift that optics is going to drive in networks. So optics is actually going to drive that shift and we're going to see a shift from what was delivered as part of an optical system to what is now delivered as a pluggable that goes into a router and that will simplify networks for customers. It's going to simplify the deployment. It's going to simplify the operations. It's going to reduce the footprint required and it's going to reduce the overall cost. So that's a very, very exciting architectural shift that we think optics is actually going to drive.

So Bill, I want to ask more of a strategic question and then drill down on technology. But from a strategic perspective, you've got a couple of acquisitions, most recently, Luxtera, and the pending acquisition of Acacia, that speak to Cisco's effort to become more vertically integrated as opposed to simply buying for merchant suppliers. Could you maybe explain the strategic thinking of why this needs to be part of your business as opposed to a partner or a supplier to Cisco?

Yes, that's a core question, Simon. So as I think many of you may know, we have a very substantial optics business today that serves all of our customer applications from service provider, web, as well as our massive enterprise customer base. And generally speaking, when we sell routers and switches, we sell the optics that go with those routers and switches. And so customers expect to be able to buy a router from Cisco and then to buy optics that work in that router. And the optics have to work in not only the 9K, but they have to work in the 6K and the -- the next generation, 8000 Series routers as well. So we take responsibility to make sure that they will work in any of the platforms that we sell and that there's interoperability. So when a customer has a data center switch at one end and they have a router at the other end, we take ownership to make sure that's going to work for them. So customers expect that, that when we sell them products and we sell them routers and switches, but with that comes an optic that's going to work. Over the past many years, we have relied on suppliers to deliver those optics to us that we qualify. We take through a very rigorous process to make sure that it works according to all the standards and specs. We make sure it complies against all the corner cases like temperature variation and voltage variation and different host environments. And we've then qualified it, put a Cisco label on it and resold it. That's been the business model for optics. As optics becomes a bigger portion of the spend, then customers become more sensitive to that cost in the network in the bomb. So clearly, you can think and recognize that a web customer is going to be looking for a very competitive price point when they're buying optics. And if we're procuring those optics from a supplier, and then trying to resell them, there's just not enough margin in the stack to be successful in selling that across all segments. And so increasingly, we believe that for some portion of that business, we will need to own that technology. It's not every optic, but for some portion of that business, we will need to own that technology in order to be able to continue delivering value for the customers and do it at a reasonable margin.

And I know that some of your peers have taken a very different strategy and the argument is that optics as a business are relatively low-margin type of business. And obviously, you haven't been breaking it out, but what makes you confident that it is a profitable business even if you vertically integrate it?

Yes. So first of all, I think you have to listen when anybody is saying that they don't think that this is an important part of the portfolio. It's generally they're going to defend that if they don't have it and say, this is why we don't think it's important. I think most of the competitors that are saying that would very much like to have that inside, if they could. And the reason is that business is a very important business for us as we serve enterprise customers, many, many hundreds of thousands of customers who expect Cisco to be able to deliver optics with the routers and switches that we're selling, same is true for service provider customers. If we're going to continue to be relevant for them, we think that the best way to do that is to basically control that cost point and not rely exclusively on third parties to do that. So I do think that it's an issue when we look ahead and say, as we look at this business evolving from 10 gig to 100 gig to 400 gig, that's increasingly technology that we need to own in order to be able to deliver that value to customers at a reasonable margin. The other thing that's important, Simon, is we do see some trends with things like onboard optics and co-packaged optics where optics and silicon come together. And at some point, in the future, that's going to be a necessary way to deliver the next capacity gains on routers and switches. So with all the massive investment we're making in silicon, we want to make sure that we can control that solution and not be beholden to others in the industry that might have exclusive or very near exclusive control over that. And so that, as we look ahead, and that, again, is not a today problem. But as we look ahead a few years, we want to make sure that we can control our destiny on routers and switches because that's our bread and butter.

So I wanted to see if maybe you can give us some update or at least refresh us on the pending acquisition of Acacia. That was a transaction announced, I think, in July of 2019, has not yet closed. So if you could help us in terms of what's the status and then remind folks of the rationale for that transaction?

Yes. So let me start with the rationale. As you may know, Acacia is an industry leader in offering coherent solutions. Again, those are solutions that are used to transmit light over long distances, anywhere from 80 kilometers to thousands of kilometers. We use Acacia in our optical portfolio today. We use Acacia in our routing portfolio, so we are a customer of Acacia's. And this is another area where we look and see vertical integration at some level makes sense for us because that DSP and the silicon photonics that wraps around that DSP to create the whole coherent function becomes a bigger part of the bomb as the bit rates are increasing. It's a much bigger part of the bomb. And so if we're going to be able to compete with others out there that may own that technology, for instance, Ciena owns that technology. Nokia owns that technology, then we believe we need to own that as well, so that was one rationale. The other rationale is that the issue that I alluded to earlier, which is that we're going to see that technology move from the classic optical system into a pluggable that is then sold as part of a router sale. And so that then becomes a meaningful portion of the network spend for customers who are deploying routers, and we want to make sure that we can capture that. That will be -- we'll view that as part of the TAM. And so we want to be able to make sure we can capture that, and we don't think we can do that unless we own that technology. We think that, that is going to be basically a multi-vendor environment. There are going to be other suppliers out there who are going to offer that technology direct to our customers and our customers are going to expect that we can offer that to them in a tested, certified, interoperable way. And if we're going to do that, we're going to have to have a competitive cost point in order to be able to lift our market price point.

And in terms of just update on the status?

Yes. So when we announced the deal, we did, I think mentioned that we have to go through significant regulatory approvals. And we anticipated that would take a year or so, though that's always a big error bar on that. We have completed the regulatory approvals in the U.S., in Austria and in Germany. And we are now in the approval process with China. And as you know, that is a -- I mean there's political considerations. There's competitive situations that we have to confront there as well as customers that are in China of Acacia. I would say that process so far has gone well. We have submitted the filing to SAMR. We've received questions from SAMR, we've responded to those questions. It's hard to predict what their views are, but I would say the questions have been questions that we -- none of the questions surprised us, and we felt that the questions were all constructive. I've had -- I've personally had discussions with each of the major Acacia customers face-to-face and have assured them that we are fully committed to this business and fully committed to making sure that this business with Acacia continues post close and that we'll honor the road maps. I think those conversations have gone reasonably well. And so -- and we've had some subsequent conversations with those customers as well. So I feel like things are roughly on track, though, given uncertainty around how China behaves it's a -- it will be -- I wouldn't want to predict a different date that this thing will close. But I would say it's on track relative to what we anticipated.

So just to be completely clear on this, when you're talking to the Acacia customers, I think there's been some concern of whether or not they'll have access to the Acacia technologies, the DSPs and the transceivers because they may, in some respects, compete with Cisco, and I believe you said that you will continue to offer Acacia products on a merchant basis even to companies that do compete with Cisco. Just want to make sure that I understand that correctly.

Yes. I mean, that's exactly right. And let me just expand on that a bit and offer a counter example. When we acquired CoreOptics, CoreOptics had several competitor customers in addition to Cisco. And we made a conscious decision at the acquisition time that we would honor the existing supply agreements with those customers for as long as they existed. But it wasn't -- we weren't interested in getting into component business based on the CoreOptics technology alone. And so at that time, we had a conversation with the customers of CoreOptics and said, look, we're going to honor this business. We're going to do that with integrity for as long as the supply agreement is in place, but we're not going to plan to renew this. We really don't have an interest in being in the component business for CoreOptics. A fundamental change that's taking place in Cisco that I think allows us to go have this conversation very credibly with customers now is we are now in the component business. We had a major announcement in December saying, we're going to offer silicon to customers who want to buy our silicon. We're going to offer optics in whatever form they want. They want to buy the pluggable optics, that's fine. If they want to buy an embedded module, if they want to buy onboard optic, which we sell today, if they want to buy chips from us, we will offer the technology in a form that the customer wants. And so that's a fundamental shift for Cisco that has taken place recently, that really allows us to say with very high credibility to these customers, we're in this business now. This is not a one-off business that we would have had to create for jet CoreOptics DSP 10 years ago. This is a business that we're building a complete business model around. And so it is really a fundamental shift from where we were 10 years ago. And I think the other point I would make is that we are looking to expand the business with these customers. We are going to have to separate the conversation with these customers from the conversations we would have internally about optical systems, for instance. But we want to expand the business with ZTE as an example, who is a big Acacia customer. And in fact, that may be the best way for us to participate in China on optics. So we're going to do everything we can to make them successful.

Great. And then you did mention earlier in your comments, this pluggable optic known as a ZR. So that's applicable in the 400-gig market. And I think you've talked about ZR as one of the aspects that made Acacia attractive to Cisco. Maybe just step back and help us understand your view of what this VR market is worth. How big a market do you see? And what's the timing?

Yes, it's a bit hard for me to answer the market size. But I would say there's a multibillion-dollar market in 100-gig coherent today. And some portion -- and whether it's DCI or whether it's a metro application, some portion of that market over time is going to move into pluggables, and we're going to see it first take place with ZR, which has a fairly limited application range of a 100-kilometer type application, 400 gig. We'll see enhancements to that ZR with things like ZR+ that expand the application range out a couple of years. And over time, we're going to continue to see that. That's where the innovation is going to take place in the industry. It's going to be to put more and more capabilities into that pluggable to drive down the power, to put more capabilities in and take more of the application space that is currently served with optical systems. So it is a shift in some sense from the optical systems TAM into the pluggable TAM. But I think it could be a multibillion-dollar shift, okay, over time. But I would expect that to begin later this year, and it's going to be very early stage with point-to-point relatively short-reach applications. And over the next 3 to 5 years, we'll see that become much more prevalent as a deployment model.

And Bill, you play in both the systems perspective as well as potentially this pluggable perspective. So net-net, you end up losing revenue as systems where the DCI application go down and pluggables go up. How should we think about the net impact?

I -- so I don't think so because we're a relatively small player in the optical market today. Worldwide -- in the U.S., we're probably like a 10% player. Worldwide, we're probably like a 5% to 6% player. Between routing and switching, we're 40% type player in the market. So if we can attach those ZR optics to routers and switches that we sell and make that part of the selling proposition and do that competitively, then we, I think, can effectively expand our optical footprint through the ZR. And so it's not a shrinking of the TAM, in my view. It's a shift of the TAM from optical into routing. And then we can expand that because we certainly have a bigger footprint in routing than we do in optical.

And what do you think of as maybe a plan B should the Acacia deal not get approved, how do you plan on tackling the ZR opportunity?

So first of all, I'm not planning on a plan B at this point. We are confident that Acacia is going to get approved. But we are -- we always have an OEM option available to us. And even in the example of something like Luxtera optic that we would develop inside the data center, we generally have 2 sources of supply. So 1 of those might be outside Cisco, 1 might be inside Cisco, and we do that for supply chain integrity reasons as much as anything. So the plan B would always be to work with other players, including Acacia and make it an OEM model.

So that makes maybe for a good pivot in terms of how you're thinking about the inside the data center optics market in that that's an opportunity where maybe Cisco's position isn't as well appreciated because of your relatively small exposure to hyperscale. But could you maybe give us a perspective of what you see going on inside the data center in terms of optics?

Yes. So obviously, there's the hyperscale market and then there's the broader enterprise data center market, an SP data center market. 100 gig is ubiquitously deployed now in a variety of flavors. You can see everything from multimode to single-mode, 2 broad categories, but you can see things like PSM4, CWDM4, LR4 and so many different flavors of 100 gig have been adopted. And that, in some sense, fragmented the industry because the supply chain is basically -- supply chain is being developers of that technology and manufacturers of it, basically had to divide themselves to say, am I going to align around CWDM4, PSM4, FR, LR or what or whatnot, or multimode. And I think that largely speaking, that -- first of all, I expect 100 gig is going to have a very long tail. That 100-gig market has certainly matured for guys like web players, many of the service providers, but has not even begun for service providers in any meaningful way -- I'm sorry, for enterprises in any meaningful way. So that is going to have a very long tail, like we still sell a ton of 1 gig, 10 gig, it's going to have a tail like that for enterprises. And I think the industry has done a very good job of driving down the cost of 100 gig. We're not quite at $1 per gig, which is what the web scale players would like to see, but it's certainly been driven down. We're introducing, actually, we would have been demoing this at OFC. We're introducing a next-generation 100-gig solution, which is a single wavelength 100 gig. Almost everything to date, everything to date in the industry has been typically 4 channels of 25 gig that put 100 gigs worth of capacity onto a fiber. And the reason you want to try to get to a single wavelength is that the laser is a big part of the bomb, the base -- the laser cost a lot. So if you have 4 of them, it's going to cost more than 1. And there have been lots of technical challenges in getting to a single wavelength of 100 gig versus 4 25 gig. But we are now introducing that. It's called the 100-gig FR, and we were introducing that, we were demoing at OFC. We'll have that in the market later this year at some very, very aggressive reaches for customers like 10-kilometer and 20-kilometer reaches, which are fantastic performance. So I do expect that we'll see some customers transition to that optic. And the reason that Optic is also attractive is that when customers move to 400 gig, there's basically 2 flavors of 400 gig that are emerging here. One is called DR4 and one is FR4. And the DR4 is you can think of it as 4 fingers, 4 parallel fingers, each 1 has 100 gig. And the question is what goes on the other end of that? Well, you can put another 400-gig DR4 at the other end of that or you could attach it to a 100-gig port and fan it out 4x. That 100-gig port needs to be the 100-gig FR that we're introducing now. So that 100-gig FR will have applications not only for customers that want to try to drive to a different cost point, but also for customers that want to fan out 400 gig in their data centers. I expect 400 gig will start to see some meaningful deployment in second half of this year. I think that the industry has been much more cautious about jumping to 400 gig, given the 100 gig experience, which was, again, very fragmented, and there were a lot of adoption pains with 100 gig that, I think, we've learned a lot of hard lessons as an industry and won't be repeated at 400 gig. And I think, frankly, the customers are more cautious as well about how quickly they're going to commit an architecture to 400 gig before they see the supply chain stabilize around 400 gig. So I kind of expect like second half, we'll start to see that pick up.

Now one of the things we've heard about the 400-gig switches is that in the case of a fully equipped switch, the optics will represent roughly 70% of the value with essentially the chassis and line cards, the other 30%, whereas it's almost the reverse in 100-gig platform. And there's been some debate as to this particular topic. And maybe if you could just weigh in on how 400-gig optics within the switch compared to the 100-gig cycle?

Yes. So the -- I mean, the rough math is what you said, Simon, the rough math, if you go back to 10 gig, the optic represented about 10% of the cost. The port was 90% of the cost. And at 100 gig, it's about 50-50. It's about 50% was in the 100-gig optic. That's come down a bit, but it's certainly moving in the direction. And it's a permanent shift. It's not like just an adoption issue or a learning curve issue. The relative cost of the optic is increasing relative to the port. And at 400 gig, it's certainly going to be more than 50%. It may be upwards of 70%, but it's certainly more than 50%, like I'm willing to stake my -- place a bet on that, that that's a fairly stable state is it will be more than 50%. And part of the reason is that the silicon guys have done a great job in driving much, much more capacity into ASICs. We've got 10 terabit ASICs now going to 25 terabit ASICs that basically, I mean, at some level, if you think about an ASIC cost is an ASIC cost. When you go from 10 to 25, the ASIC cost doesn't go 2x or 2.5x. And so they've driven down the cost per bit by cramming more and more into the ASIC, thanks to Moore's Law. And then so they've driven down that effective cost per bit, the optics costs have not come down as fast. They've come down, but they have not come down as fast. And the reason is that optics isn't just an ASIC. It's got all this other stuff like lasers and discrete components and fiber coupling that you have to do in order to get the optic to work. So it's a more complicated development, manufacturing and assembly process. And so those costs have come down, and they will continue to come down with things like silicon photonics, but they're not coming down at the same rate as the port costs. And so relatively speaking, as you go from 100 gig to 400 gig to 800 gig, the relative cost of the optic is increasing compared with the port cost.

And do you believe that silicon photonics will really make that big a difference given that ahead of the OFC or around OFC, we've seen a number of announcements from traditional transceiver makers offering 400-gig platforms that are not using the same kind of technology that Cisco intends to use or has used. Is there a meaningful change in terms of features, function or even cost in terms of the design if it's a standardized pluggable?

So first of all, I don't want to come across as religious about silicon photonics. It's an area that we think is -- has a lot of headroom for improvement in the cost structure and in things like how much can you bring into the silicon photonics ASIC. But if somebody came along tomorrow and said, look, we've got an Indian phosphide solution that's half the cost of silicon photonics, we would jump on that immediately. We're not going to be really just about the technology. We do believe that silicon photonics has a -- is on a trend to get to better cost points and better power points. And so I think long term, initially, we're going to see 400-gig solutions out there that use very conventional technologies. And I think longer term, we're certainly working to integrate more and more of the function into the ASICs, the photonic ASIC and the electrical ASIC in order to minimize the bomb effectively, minimize the number of parts you have in the bomb. And silicon photonics offers that promise to be able to do that. And so I think longer term, it has an opportunity to get to a lower cost point and to a lower power.

So I wanted to pivot to the more traditional systems business. So it's well known that you've secured a piece of business with the metro project, Verizon, that you share with one of your competitors. And when we looked at this optical systems market, it seems to be consolidating among the biggest 3 players, and you're not among the biggest 3 players, unfortunately. So how do you think about your fit strategically and competitively in what has been the more traditional business for optical systems, primarily selling to service providers?

Yes. So optical systems for Cisco is a -- I would say, as we look at the portfolio for service provider, we obviously have edge routing and access routing and core routing and cable and optical and automation and mobility. And optical helps fill up that portfolio. But Cisco is not going to stand up and say, we want to be #1 in optical. That's not a goal of Cisco. We'll play for optical opportunities, as they help us expand the overall opportunity for Cisco. So as an example of PLDT, PLDT was a huge win for us in the Philippines, a nationwide network build out. PLDT said, I want an IP plus optical solution. I'm going to buy an IP plus optical solution. So you need to bring both. And that was a perfect opportunity for Cisco to put together both and say, we can deliver you an IP optical solution that a stand-alone optical vendor can't deliver and a stand-alone router vendor can't deliver. In the case of Verizon, I think we had early -- and Ron Johnson is on the call here as well. We had early and intimate engagement with them where we actually influence their architecture. And I think we built a lot of trust with them. And fundamentally, I think it came down to they placed a bet on somebody as they were going to trust would be there for the long haul, and we'll be continuing to invest for the long haul in optical as well as routing. And if you remember, for Verizon, it included things like we have a packet optical solution. Verizon started out with basically looking at an OTN architecture, saying, we just want to build an OTN architecture. And we work with them very deeply in looking at their service needs and their expansion, their -- what they anticipated the network seeing over the next few years. And we jointly came to the conclusion that OTN architecture would not scale. It was the wrong answer for them. Basically, next-generation SONET was just the wrong answer. And that they really needed to be betting on a packet-based solution. And then that led to a conversation around, what I do with my TDM services? My TDM services have to be carried on an OTN or a SONET infrastructure today. And we came up with an architecture to migrate their TDM services, whether it's OTN or T1 or an E3 to a packet-based infrastructure, TDM to IP. And so we worked very deeply with Verizon on a number of architectural challenges, but there was a linkage to the packet and routing portfolio as well. It was not purely optical. Now it's turned out to be a massive, massive deployment for us for optical. But it's basically, we built credibility with them over time by looking at the architecture very holistically and not getting sort of pigeonholed into it's got to look like an OTN solution or a DWDM only solution because that's all I have in my portfolio.

And did you have any OFC announcements related to the systems side of the business?

I think most of the OFC announcements were demonstrations of things like ZR technology, for instance, of some of our 400-gig portfolio. I think it was more along the -- probably the ZR was where we had most of the -- most of the announcements.

Okay. So why don't we pause for a moment and ask the operator if we can poll for questions from folks dialed in on the call.

[Operator Instructions]

While the operator is checking, Bill, I've got somebody e-mailed me in a question actually. So we've heard in terms of the systems market, a number of companies talking about launching 800-gig platforms and a number of products launched with 600 gig. So there's sort of this one-upmanship out there. What's sort of Cisco's current state of the art? And how do you think about competing in that kind of race for faster speeds?

Yes. So thanks for asking that. I think there have been a number of announcements out there. You guys can probably name the guys that have done that. I don't -- I have not seen anything deployed, frankly. I think a lot of these guys have been very aggressive in their announcements and their claims, for instance, deploying by fourth quarter of last year. I don't think anything has been deployed for 800 gig. And in fact, I don't expect to see that deployed anytime very soon from what I'm hearing from customers. I have not even seen it in customer labs, to be honest. But I think the bigger point here is we are deploying 600-gig solutions today. Those are in mass deployment. And we've been deploying that for about a year now. I think as you look at the network applications, there is this trade-off that always exists between capacity and distance. So as the capacity goes up from 100 to 200 to 400 to 800, the distance generally gets shorter. You can play some tricks to try to overcome that. But at some point you run it's a fundamental limit of physics. And so the real question is what's the application space for 600 gig or 800 gig, and it's a relatively small piece of the market. Like that is much more about bragging rights, in my view, than it is about meaningful applications. The sweet spot for the network is going to remain 200 gig, 300 gig, 400 gig. Those are going to be the real sweet spot for, like, a long-haul network or a metro regional network. You will see 800-gig applications or 600 gig in very short distance applications. But the big deployments of long-haul metro are going to require longer distances generally, and you will see 200, 300, 400 gig. The other thing -- so I'm not -- I don't think chasing -- like for us, it's not an objective to say, well, we need to come out with 1.6 terabits. That's not an objective. A much more important objective for us is to say, I think there's going to be a very significant shift to pluggables and routers with ZR kind of leading that charge, and then watching that and driving that innovation over time to drive down the cost for a pluggable and to drive up the application that it can serve, meaning longer distances. And so my focus will be on trying to expand the application domain for a 400-gig pluggable in a router and to drive down the cost of that as opposed to going to chase like can we -- if somebody says they can do 800 gig, can I do 1.2 terabits to 1.6 terabit, that's not the game we're chasing. Ron, you are on. Do you want to add anything to that?

Yes. I think the other dynamic that's important to think about here is that we're approaching the channel limit. And these technologies are not -- as we incrementally increase the bit rate and play games with a BAUD rate, we may be able to get away with fewer wavelengths to fill the C-band, but we're not necessarily getting more capacity into the C-band. And so cost, in this case, becomes a very important factor, right? And if it's significantly more to increase the BAUD rate or change the bit rate to address this, it doesn't necessarily mean that it yields a lower-cost solution. The lower-cost solution could be yielded by more wavelengths at a commoditized price point or at a price point that has a very strong component base and integrated components.

Well, great. So as we're coming up on the top of the hour, I want to thank Bill and Ron, and Emily for joining us. Bill, I'd like to just give you an opportunity maybe in closing to sort of soundbite this. And what I mean is, so optics obviously an important part of Cisco's business, missed an opportunity to meet and hear from you at the OFC trade show. So given the nature of, let's say, a financial analyst audience, if there was some soundbite you wanted to make sure people left this call with, how would you like to wrap it up?

I would say that, first of all, thank you all for your time today. I appreciate that very much. And I would say that optics -- keep your eye on optics because I think optics can actually drive architectural transitions in networks, whether it's inside a data center or outside the data center as opposed to just follow those transitions. That, I think, is a very key point for how we're thinking about optics.

Well, great. Well, folks, thank you very much for joining us. We're just coming up on the hour. Sorry about the challenge some had dialing in. We hope to address that in the future calls. And Bill, Ron, Emily, thanks for joining us. I hope everybody has a good safe weekend. Thanks.

Thanks, Simon. Thanks, everybody.

This concludes today's conference call. Thank you all for joining. You may now disconnect.