The following is a transcript of the panel conversation between Grant Joslin, Telecom Equity Research VP at Credit Suisse, and Roy Chua, Principal and Founder, AvidThink, during Credit Suisse’s 2022 24th Annual Communications Conference in June 2022. The content has been lightly edited for readability. The on-demand recording of the panel is available for viewing (duration: 30 mins).
Grant Joslin: Good afternoon, everybody. Our next session this afternoon is going to be on telco cloud transformation and 5G’s enterprise future, and our speaker joining us is Roy Chua from AvidThink. Roy, thanks so much for being here.
Roy Chua: Glad to be here, Grant.
Grant Joslin: So for those in our audience who aren’t familiar or missed your presentation last year, can you introduce yourself and AvidThink and talk a little bit about what you do day to day?
Roy Chua: Absolutely. So my name is Roy Chua. I’m the principal analyst here at AvidThink, and AvidThink is a research and analysis organization. We used to be part of a larger media and research company I co-founded in 2012. But in 2018, we spun it out as an independent analyst firm. What we focus on primarily is infrastructure technology. So think networking, cloud and edge infrastructure, infrastructure security. Primarily we provide research and advisory services to some of the tier-one service providers worldwide, hyperscalers like AWS, infrastructure technology vendors, Dell, HPE, VMware, IBM, folks like that, as well as enterprises, private equity, and management consulting firms. So that’s the type of work we do.
Grant Joslin: That’s a big list. So it’s been a year since we last sat down to talk about telco cloud and 5G enterprise at the 2021 communications conference. So in the last year, where have your views changed the most, or what surprised you?
Roy Chua: It’s been an interesting year. It’s sort of half out of COVID and half not. And what I would say is I think what I’m encouraged by is this whole hyperscaler-telco interaction. I think that’s something that’s hard to ignore, because it’s a meeting of the giants. Right? You got elephants dancing around, and so what do you do? I would say I’m encouraged by the ongoing collaboration. I think 18, 24 months ago, people, well, we included, were writing about the analysis between what was going on, and some of the folks that we work with, both on the hyperscaler side and the tier-one telco side, were trying to figure out how to make this work.
Roy Chua: And so it’s kind of weird where on one hand some of the clients are telling you, “Hey. What do we do with the hyperscalers?” and then the hyperscaler asking you, “So what do we do with a telcos?” Right? But that’s no longer the case. Now I think the sentiment has changed. There’s a lot of more collaboration as we see going forward, and I think, in general, all the tier ones are kind of tied up with all the major clouds, and all the major clouds are kind of tied with all the tier-one telco.I think that’s going to be the reality. So I think that’s a good thing.
Roy Chua: Other than that, I would say that the private wireless continues to see traction, and I’m sure we’re going to talk about that later, but we’re seeing more traction. It’s still limited. The supply chain constraints. Some of the ecosystem pieces are still not quite there yet. But despite that, it’s still getting good traction, and so I think that’s the good one. And CBRS continues to roll out. There is the complexity on the device side. I think it’s not as seamless or as easy as it really ought to be. But the AWS private 5G rollout certainly surprised a lot of people in the market, and the speed of the rollout surprised us as well, and there’s very strong demand. So clearly the markets are looking for something like that.
Roy Chua: And then finally, more recently, I think what surprised me was Broadcom bidding for VMware, I guess. Good and bad, depending on which side you’re on.
Roy Chua: So, regardless, it’s going to change the shape of the whole software infrastructure layer ecosystem. So we’ll see what happens there.
Grant Joslin: There’s a lot there we’ll dig into over the course of the next half hour. To start with telco cloud, I think any discussion of telco cloud has to feature Dish at some point. Looking in from the outside, how would you rate Dish’s build so far? Is it profound problems? Or is it just the normal amount of work to orchestrate and coordinate everything? Or is it actually smoother than you expected?
Roy Chua: I think that’s a very, very timely question, actually, given the press release that Dish put out this morning about them actually meeting the 20% commitments. Right? So 120 cities. I checked right away. I’m like, “Ah. Not in my city yet,” on the Project Genesis side. But good. I’d say that’s kudos to them, because I would say it’s probably surprising in terms of how smooth it was in the grand scheme of things. I know that it’s a little delayed, a little slower than I think we were all hoping for. But given the number of vendors … I mean, to be frank, I mean, they have a lot of vendors on the ecosystem list.
Roy Chua: On the radio side, you got Fujitsu, MTI, and now Samsung. We got Mavenir, Altiostar, Cisco, Samsung, Nokia. I mean, the list goes on.
Roy Chua: And if you look at all of that, how do you coordinate? I mean, and it’s one thing where we look at Rakuten, which sort of got there first, they had a more constrained set of vendors. Some of it was customized to their needs. But what Dish has done and the team over there has done is they’ve taken a lot of vendors and saying, “I’m going to make this all happen,” and then integrating all of it. That’s non-trivial. I mean, the reality is it’s non-trivial. I’m still surprised that they actually made it and they have a network that’s in production and all lit up. That, I would say, is actually kudos to them, and they’ve been smart about how to manage and leverage those vendor relationships and orchestrate those vendors.
Roy Chua: I mean, part of that, staying under the 10 billion CapEx line, which is still holding, I think, is leveraging a lot of the efforts on that vendor side, because who doesn’t want to be involved in the first greenfield, cloud-native, open multi-vendor network in the US? And so we work with some of their vendors. Some of them are clients, and it’s a surprising amount of resources they’re willing to put in to support Dish. And so I think they’ve benefited from that, for sure.
Grant Joslin: So Dish’s VP, Sidd Chenumolu, said last month, “We will change vendors. We have to adapt to the growing industry needs. Customers are our priority. Not the vendors.” So was he talking about Amazon? Is AWS a smaller piece of Dish’s network than we originally expected?
Roy Chua: So I would say that from the early days in terms … I mean, number one, I’m not sure who Sidd exactly is talking about. Obviously, he hasn’t shared more details since, and I don’t think he’s going to. I do know that Dish has been very smart in keeping all the vendors on their toes. That’s for sure, in working with some of their vendors. What I would say is this, is that I think AWS, when it was first announced, that relationship, that they were core to Dish. I think they still remain core to Dish, and speaking of core, they run Dish’s mobile core.
Roy Chua: So it is critical in their network. I think there were some elements on more of the edge side of things, where AWS with their outposts and their edge offerings, I think, were expected to play a somewhat bigger role. I think that remains to be seen. I don’t think we’ve seen all the details of that RAN network itself. We do know that what’s been consistent so far is Marc Rouanne at Dish talking about the use of Dell and VMware, and obviously Mavenir software at the edge. And so I think that part’s been consistent.
Roy Chua: Now, the mix of Dell versus AWS Outposts, I think, remains to be seen over time. But in terms of the core elements and the workload-wise, I think AWS is still their partner for that. And the reality is, from Dish’s standpoint, is that – and Marc’s been very clear – is what they own is the workflow. What they own is the pipeline. His view is that the infrastructure layer underneath that, which is AWS and Dell and even VMware … Those, he views as switchable. Right? Those, for him, are fungible, and he’s happy to switch those.
Roy Chua: There are switching costs. I mean, I think that’s the reality yet. When he says it, it sounds simplistic. It’s not. It’s actually quite complicated, but he’s designed it, at least in his mind, so that those can be moved around if and when needed, and I suspect they will probably make some of those moves strategically to continue to get the price advantage and to keep the vendors on their toes. I think that’s been the way they’ve been managing vendor relationships.
Grant Joslin: Another question that we have gotten is whether a legacy traditional RAN vendor like a Nokia or an Ericsson or a Samsung being in the mix for this build loses any of the benefits that they’ve talked about with cloud RAN and open RAN.
Roy Chua: So what I would say is that the reality is when we talk about the legacy vendors, the Ericssons and the Nokias, they have been undergoing transformations themselves, and the parts of the solutions that are being adopted, like, for instance, Nokia is part of the Dish Network … Those parts are in the process of or have completed the cloud-native transition, and so the elements that they’re pulling from these vendors are the cloud-native versions of those vendors. And, I think, in the recent investor day, I think Marc was on stage talking about their principles, and he’s been quite stringent about the adherence to those principles, and part of those are the cloud-native principles. Right? It’s got to be zero trust. It’s got to be cloud native. It’s got to be blah, blah, blah. Otherwise, you’re not allowed on the network.
Roy Chua: So, I think, to the extent that the incumbents are involved, it’s the more modern version of the solutions that incumbents provide. They’re not buying into the legacy stacks. They’re taking the newer things and pushing them using their position as Dish to push those incumbents into bringing the newer solutions and re-architected solutions into that network. So I don’t think they’re held back, per se. In fact, what I’ve seen as far as I’ve seen of the Dish pipeline and workload and orchestration is, in fact, it’s definitely a modern architecture compared to some of the more legacy MNOs out there.
Grant Joslin: Got it. So we talked a little bit last year about Dish chairman Charlie Ergen’s prediction that Dish would be AWS’s largest customer someday, and Dish is just one mid-sized network in one country. So when you scale that up to the whole world’s telecom networks, it would seem to suggest a huge opportunity. So are the hyperscalers investing, like wireless or telecom is, this giant opportunity?
Roy Chua: They are. Based on the size of the teams that they’re building up across all the major hyperscalers, my belief is that they are. They’re actively recruiting. I know a lot of people that are working there. So they do view it as a significant opportunity. With regard as to where they are, I think the back office operations … There’s little doubt in my mind that many of the telcos are migrating to the public clouds or have already, and so those workloads are in process of moving there. So it’s definitely a huge opportunity for all the analytics, for all the data. The core, the mobile core especially, and even the wireline core … I think that’s in process of gaining. AT&T is tied up with Microsoft Azure recently to bring the 5G core in there. Obviously, Dish is running there. I think we’re seeing some of that. The RAN, the things that are out further, or the edge … I think that’s in progress. We’ll see what happens with that on the RAN side.
Roy Chua: And then on the enterprise side, where we all view big dollar signs, Dish and other guys included, on the private 5G, the private network side of the equation, that remains to be seen. I think the hyperscalers understand that that’s sort of where they may be weakest in some sense from a presence and proximity standpoint. But they do have one advantage, which is the dev pipeline. Developers … They hold the developer application relationship. So that play, I think, is ongoing. But hyperscalers recognize the value of that. All that connectivity, they want … They don’t really care about the connectivity as much. They just want the data to come to them, at the end of the day. Because that’s where they make the margins. So we’ll see. We’ll see what happens. But I think the wireless companies will become large customers, very large customers of the public clouds. I think that will be a new trend.
Grant Joslin: So it’s interesting that you mentioned AT&T and Microsoft. Frank whom I spoke with this morning has argued in the past that telcos are outsourcing their brains with moves like AT&T selling its engineers and cloud tech to Microsoft. Do you agree? Or maybe more broadly, what’s the right amount of technology work to do inside of the company versus to buy from a partner?
Roy Chua: I think the answer is that there is no one answer. I think, cynically, you could argue that a lot of the brains or the software to run the brains are being developed by the hyperscalers with regard to analytics, AI, AIOps, all those things. So in some sense, it’s not outsourcing their brains. It’s basically borrowing better brains for some of the operations in the telco. So I think it’s how you look at it.
Roy Chua: I mean, it’s leveraging, and I think the question there is where is your leverage and how do you retain control. Right? And so the telcos that we work with right now, like NetCo, InfraCo, ServCo, … I want to be a TechCo. But I don’t think they all understand how to get there yet. That’s still an in-progress thing. I mean, cynically, what we don’t want to happen fundamentally is for telco to become basically a marketing engine with spectrum assets and the ability to do regulatory compliance. Right? And that’s all they are. I think they want to be more than that.
Roy Chua: So I would say that I don’t think that AT&T selling that portion to Microsoft makes their brains go away. I think, in that sense, they’re borrowing better abilities outside the organizations and leveraging best-of-breed outside. The danger with doing that is I don’t know if they fully understand what the new business is. And I know for Dish and Marc Rouanne it’s the workflow, it’s the automation, it’s the API element, which is what he’s trying to retain as part of the business. Will that be sufficient? Will that be sufficiently valuable in terms of maintaining value to the investors? Remains to be seen.
Grant Joslin: Got it. And then I think one other place I’d really love to get your input in is cost, because cost reporting, for these big wireless companies especially … They’re just kind of one big bucket of OpEx, and we never really get to see much more than that. So at the same time, AT&T and Verizon and all these guys constantly have efficiency and savings initiatives going. So are there a lot of cost savings on the table from telco cloud, and are a lot of those yet to be demonstrated? Or are they already saving a lot of money from this move?
Roy Chua: So the telcos that we work with or the ones that we’ve seen some of the numbers in there … I mean, consistently by going to a cloud-native … And cloud-natives are two elements. There’s the cloud-native part of the infrastructure, the CapEx side, and then obviously the cloud-native automation-driven, so the OpEx sides. And the numbers we’re seeing, generally speaking, run between 30 to 50 percent savings, I mean, across the board, and that’s not surprising. When we look at the web scalers, the web hyperscalers, and the web applications, that kind of savings were already there. So by changing software development processes, automation, 30 to 40 percent we’re seeing sort of across the board relatively consistent. Whether they’re sustainable in the long term depends. It depends. Right? And can you get better than that? I think that also depends, but it is possible. 30 to 40 should be achievable. We’re seeing enough numbers now that that is an achievable goal for a reasonable telco.
Grant Joslin: Got it. So when we take this trend and we fast forward a couple years and we’re sitting here at the 2026 communications conference or something, how are telecom companies, tech organizations going to look different? You know? Are they going to be like thinner organizations? Are they going to be more like systems integrators? Or can you make a wild prediction about that for me?
Roy Chua: Yeah. I don’t know if it will happen by then, but I do see a couple of scenarios or elements play out in maybe all of these kinds of animals. Right? So the telco … As you will. Right? Types of or categories of animals. I think, on one hand, I would say that there are sort of the thin operations that are basically there and they’ve optimized, so kind of like Dell for the servers. Right? So I’ve optimized my business, I make small margins, but I’m highly efficient, and you’re not going to beat me at my business, and I have API, I have platform operations, and I make it available. You run your businesses of services on top of it. I give an API. I’m low cost. I’m decent margins. But that’s all.
Roy Chua: I think, on the other side you have the result of the Rakuten type approaches, which is that I actually have a different business that I can use to monetize, and telco or the mobile is a way where I get to monetize the business or I create synergies and I make my money elsewhere. So I think you’re going to see some of those elements that come in, and the telco communications portion is just an element of that.
Roy Chua: And then the other ones, I would say that … What happens to the rest? I think some of those will be system-integrated-type businesses, because that’s where the margin is. A lot of the Asian-type telcos who have strong SI businesses will go down that route, because that’s where the margins are. And increasingly, I think the communications piece of things will reduce in value, and the business services, the business applications, the analytics of the data, I think, will gain. I think that’s just going to be reality.
Roy Chua: And in some cases, I don’t know if it’s a race to zero, but it’s a race to pretty low, from a pure, simple … Just communications. And I know the telcos are trying to fight back by saying, “Look. I’m going to charge for SLAs. Just end-to-end QoS, I’m going to charge for.” Look, we’re a couple of years now into 5G buildouts, and there are limited situations where you can charge at premium and maintain that premium over time. And what’s happened to those premiums is they’ve eroded. Look at MPLS versus SD-WAN. You know?
Roy Chua: That erosion is already happening. Right? I think that’s going to be a reality.
Grant Joslin: Got it. So I would love to check in on that next year and as we go forward-
Roy Chua: Let’s see what happens.
Grant Joslin: … and we’ll see how those predictions pan out.
Roy Chua: Exactly.
Grant Joslin: So the other area I really want to spend some time on is these 5G enterprise applications, and you already spoke about private wireless a little bit, so let’s start there. So, I think, last year we spoke about how industrial ethernet or enterprise Wi-Fi markets might be good analogs for private wireless. Has your thinking about the total market size or carriers revenue models changed with the momentum you’ve seen this year?
Roy Chua: Yeah. So I would say that I think the total market size, I think, remains the same, I think. The predictions range from between 10 to 18 or 20 billion dollars, and some are wildly bigger than that, which I don’t believe. Right? I mean, the enterprise Wi-Fi market runs in the six- to eight-billion-dollar range, but that’s just connectivity at some value. And on the top of that, you get to a 10-billion-dollar number, which is believable.
Roy Chua: I think that size is achievable. $18B? $20B? Possible. $50B? $100B? I think those are not viable numbers. Right? I would say that I still believe that private wireless is important, and I’m seeing that even with the rise of Wi-Fi 6E and Wi-Fi 7 coming in. I think there’s still a lot of strong interest in 3GPP, 5G-based technology. So I think that makes sense.
Roy Chua: So I think total market size … Still the same as last year. The carriers’ revenue models … I think that they’re struggling, in terms of trying to attack the market, and I think the reality is the monetization from that will come from the SIs will come from the network equipment providers. I think, depending on the region, some telcos will monetize and some telcos won’t, but some of them are a challenge.
Roy Chua: I mean, I think they’ve realized that the custom, “Let me build it for you. Let me charge you a huge premium,” doesn’t make sense. I mean, I’ve seen some of the quotes, having worked on some of these RFPs and RFQs. The quotes that the telcos come in with, the Nokias and Ericssons and all that, are pretty high, but the market’s not responding to that, per se. The fact that Verizon had to go with Celona for what they call low-end opportunities … Well, what happens with low-end opportunities is that they eat the market from the bottom up.
Roy Chua: So the fact that Verizon had to partner with someone like Celona, which is turnkey enterprise Wi-Fi model, I think, says to me that the carrier’s value-add piece … It’s not clear to me that they’re going to be able to charge for that in the long term and that the market really wants a Wi-Fi-type experience with better quality of service, with better privacy and security, which is what they actually want. They don’t want a custom network. They want a network that works that’s low price and better than the Wi-Fi one that they have right now. Right? Yeah. So I think the custom services that go into a private Wi-Fi for carriers … I think there’s going to be a bit of struggle with the exception of very large corporations. Right?
Grant Joslin: And I think you said monetization through systems integrators or network equipment before? So does that mean you go out to Accenture or somebody for the network and then Accenture uses Verizon just to get spectrum and a couple pieces of it like that?
Roy Chua: That’s correct. All the telco – maybe not outside the US market – has an SI consulting service in there, and they go in there and they build you a vertical application solutions for, I don’t know, agriculture or industrial IoT, where they have integration services in the carrier itself, and they monetize it. In the US, it’s more likely to be a partnership, and then the SIs will probably … In that relationship, I suspect the dollars will go to the SIs more so than the carrier, I suspect.
Grant Joslin: I think, when we spoke last year there wasn’t really either available equipment or a really pressing need to do a private network in 5G instead of in LTE. Is that still the case? Or are we seeing a lot more 5G?
Roy Chua: I think it’s still the case. It’s still the case. I know that for the lower cost, easier deployment models for private networks, the equipment for 5G is just being tested in the labs right now for some of the clients that we work with, and so a lot of it is still 4G out there that’s being deployed. I think, in a year or two, I think that will flip. I mean, supply chain constraints are slowing things down a little bit.
Roy Chua: The other element too is that with a lot of the early use cases that 4G LTE actually meets a lot of the needs. And the struggle is not so much whether it meets a speed or not. The struggle is more around how do we get compatible devices, how do we orchestrate it, how do we roll it out, how do we manage it. I mean, we’re still in that, and it’s better, but we’re still stuck there right now. It’s not like, “Oh. I really need a 10 gig or a five gig that millimeter wave gives me.” It’s more like, “How do I roll this out, how do I integrate this, how do I ensure security, and how do I get the traffic back in my local area network?” because this is basically … It runs back to the mobile core. That’s not the architecture that I as an enterprise IT person am familiar with. Right? Those are the things that I still see as being sort of the challenges in terms of deployment adoption. Right?
Grant Joslin: Okay. And then what about on-premises edge? So that is kind of joined at the hip with private wireless networks. Right? So are the carriers similarly not as well positioned for all the bespoke work here? Or are they doing better in that part of the market?
Roy Chua: I would say that they’re still piloting. They’re still trying to find traction. It’s still a challenge for the carriers to go in on on-premises edge deployments with the enterprise. I think that’s what’s happening, because in order for the carriers to come in, there has to be a carrier value-add component to that, and the carrier component … Generally speaking, you could argue that’s private 4G or private 5G. Again, it’s not clear that the adoption favors the carriers yet at this point. I haven’t seen a breakthrough business model yet for the carriers, particularly in the US.
Roy Chua: I mean, China is different. There isn’t CBRS. Right? It’s tied to it. Right? Huawei goes out, just deploys it, China Mobile’s involved. Right? That’s just the way it is. In the US market, it doesn’t look that way. Right? The dynamics are such that it doesn’t look that way. So there’s still a bit of struggle. And I think, for that one, it remains to be seen. Hyperscalers still have strength on even on-premises edge. They’re pushing very, very hard, and we’ll see what happens in that market. It’s not clear that carriers have an advantage there.
Grant Joslin: So let’s switch over from on-premises to the more important and more controversial flavor of edge compute, so public network or multi-access edge compute. Are you any more or less sure than last year that there’s real demand for that product? Has your thinking on this evolved at all?
Roy Chua: We got involved very, very early. In fact, we have five papers that AWS commissioned with MEC. And that’s sitting on the website. You know? Wavelength. Because AWS was one of the first ones there. In fact, AWS is still the only commercial MEC offering out there. Right? AT&T’s got trials going, but AWS is the only one going out there with a mobile edge. So it is happening. So I’ll say I think it’s happening. There is demand. It’s still early. We’re still in the pilot phase. I think that’s just the reality.
Roy Chua: So I think one of the key workloads that will unlock this is mobile-based extended reality, mixed reality, virtual reality, and that needs some time to develop from an application standpoint, from a headset standpoint. The other one that would drive it is the C-V2X, the vehicle to anything, cellular-vehicle-to-anything connectivity. But in order for that to happen, you need actually ubiquitous coverage for the automotive manufacturers to come in. And plus, with the model years being designed, there’s a design win. It’s a couple of years to get to market. So I think it will happen. I think it’s still a couple of years out for this mobile edge thing.
Roy Chua: I would say that the silver lining in what I’m seeing is not so much the mobile edge element, per se. It’s the local edge. So the local edge, on the other hand, is gaining traction, and that’s evidenced by AWS suddenly deciding to roll out a lot of local zones. And what the local edge gives you is it gives you the latency advantage, but it gives you the benefit of having both wireline and wireless edge that’s carrier agnostic. Right?
Roy Chua: And so a lot of enterprises that we work with that we consult with … They say, “Roy. Edge strategy. What do you recommend?” Frankly, we tell them, “Do local edge,” because it gives you the software re-architecture. It gives you the benefits of having a distributed edge. And by the way, when you deploy it, you get both wireline and wireless connectivity to your local edge. It turns out from a latency standpoint – we’ve measured this – is the latency over a mobile network to a local edge is actually pretty good. We’re talking only a couple of milliseconds difference from a mobile edge, at least today. When 5G SA rolls out and all those things happen, then obviously the mobile edge is better. But right now, as a starting point, that local edge, on the other hand, is turning out to be actually a very nice offering.
Grant Joslin: So that sounds like good news for the hyperscalers, but I’m not sure if it’s good news for telecom companies. Right? Is there a place for Verizon or Comcast in the world where you do your edge deployments on AWS local zones?
Roy Chua: There may be, because remember a lot of these guys, like Verizon, AT&T for instance … They have wireline operations, and they have wireless operations, and they have real estate with the COs, the central offices or next-gen central offices sitting there, the SAPs and the MSCs and MSOs. So for the carriers that have those unified assets and the real estate, there is no reason why they couldn’t find a way to take advantage of the proximity as well. Right? They just have to figure out that whole DevOps pipeline and their telco cloud and making it available to customers and developers in a way that is easily consumable.
Roy Chua: So I don’t think we’re done yet. And in fact, from our standpoint, we have something called the new middle mile that we’re looking at in great detail, this whole new middle mile where you have the colos, you have the telcos, your hyperscalers. I think that whole market is changing. It’s not just connectivity. It’s connectivity and compute. And so that’s, for us, a very interesting space that we’re looking at in great detail over the course of the next year.
Grant Joslin: Got it. And then I think the last enterprise application that’s always in the trio of buzzwords is network slicing, and, I think, when we spoke last year we thought there was one network slice out in the world, so very early. Have things changed now? Is it maturing?
Roy Chua: It’s maturing. Things haven’t changed dramatically yet. So you can’t buy a network slice from AT&T or Verizon or T-Mobile in the US yet. China continues to be the one place where you can buy network slices. Again, this end-to-end SLA is still hard to achieve. We’re in the process of trying to get there. We’re not all the way there yet. There’s a lot more complexity in that. And we work with all the orchestration vendors as well. Right? Your Netcracker, your Amdocs, and all next generation ones. It’s a process. It’s ongoing. I don’t think we’re there yet. We’re not there yet. Maybe ask me next year, and then we’ll see. Maybe.
Grant Joslin: Yeah. We’ll see. Exactly. Keep checking back in on it.
Roy Chua: Maybe it’ll change by then.
Grant Joslin: How many of the cool enterprise 5G opportunities require you to have 5G SA? Because it’s been really interesting for me to see that the two carriers in the US with really big business services are both 5G non-standalone still, and then T-Mobile standalone, but obviously not as big in business.
Roy Chua: Yeah. I would say that aside from the mixed reality workloads, and that will obviously require MEC presence, because otherwise why do you care about 5G SA, a lot of the workloads today, that ultra-low-latency stuff … The demand isn’t quite there yet. It’s a chicken-egg problem. The applications aren’t quite there yet. The consumer demand or the business demand isn’t quite there yet. And in fact, if you look at millisecond count, even with 5G SAs, T-Mobile cores, we’re still seeing 20, 30 millisecond round-trip times when we measure it. And so Verizon and AT&T will range from 30 to 40 to 50, sometimes 60. But the workloads that we’re still running for enterprises don’t require that. In fact, the next generation workloads for XR … You gotta get down in the sub 20. Right? And sometimes they want 15 ms for control applications. The carriers aren’t there. The market’s not quite there yet. And in those cases, it’ll probably go to private for the time being or be wirelined to a local edge for the time being. It will change. It will change. We know it’s coming. Not there yet. That’s all.
Grant Joslin: Got it. So this has been a really interesting half hour. I think it’s both been interesting to see how much has changed in a year and how many things are still kind of too early to be commercial or wait-and-see stories. So thanks for taking us through all of this today, Roy.
Roy Chua: You’re very welcome, Grant, and thank you for the opportunity.