Tantra’s Mantra Podcast – Episode 48

Prakash Sangam

Hello, everyone. Welcome back to another episode of Tantra’s Mantra, where we go behind and beyond the tech news headlines. I am your host, Prakash Sangam, founder and principal at Tantra Analyst.

Well, this is the analyst event season now, where companies are looking to review what they did this year, and give a vision for what the next year and beyond holds. We attended another one of those analyst events a few days ago. That one was from Ericsson, held at their D-15 Labs Innovation Center in Santa Clara.

The day was jam-packed with lots of good sessions. Most importantly, they had presentations and panel discussions with many of their partners, of course. There were reps from the big three, Verizon, AT&T, and T-Mobile, as well as from Meta, Intel, Google, AWS, Dell, and more.
Didn’t see anybody from Microsoft though, so that’s interesting.

The theme of the event was high-performance programmable networks. I think, of course, the emphasis was more on programmability of the networks, as high performance is always given.
The focus was more on monetizing 5G and not technology evolution itself. 5G advanced was only mentioned in the beginning of the day.

All the discussions were around APIs, AI, enterprise connectivity, and so on, which I think makes perfect sense. To discuss and analyze all of this, I once again have my friend and fellow analyst, Leonard Lee of Next Curve with me today. He’s been a frequent guest here at the show, especially discussing the event, as I mentioned in the beginning, this is the endless event season, so he’s here quite often.

Hey, Leonard, welcome to the show.

Leonard Lee

Thanks for having me on again. It seems like I’m on here all the time. Of course, yeah, and me on yours, right? Yeah, and it’s going to be interesting to have this discussion with you. They presented a lot of stuff, and I look forward to this discussion, and let’s get it started, why don’t we?

Prakash Sangam

Absolutely. Yeah, let’s start with that. What was your overall impression of the show, or rather the event?

Leonard Lee

The event, I thought it was really well done. It was, like you said, a lot to digest. I think there was an intent to provide a very comprehensive update as to what Ericsson is driving with what they’re calling high-performance programmable networks. But then also building some context around the why, around GMP and the new co-announcement that they made earlier this year.

So it was just a couple of months ago that they did in Boston. So, but I was really surprised by the guests that they brought on, that were part of the broad ecosystem that they’re putting together, especially around the GMP, you know, new co-effort. And yeah, that was pretty impressive. And I thought the many, many, many partners that they brought together at this event added tremendously to the discussion.

And yeah, I thought it was a very special occasion. And, you know, exclusive, I thought it was pretty exclusive for what ended up being a pretty, you know, exclusive audience of analysts. Yeah, there really is, you know, a select audience. So select analysts, which is good.

Prakash Sangam

And I agree with you, just jam-packed, right? It was, you know, before we could take notes and, I was live tweeting and you, I think we were on LinkedIn before we could finish what we learned in the previous session. You know, there are already a lot of interesting discussion happening in the next session.

Leonard Lee

So, you know, my fingers were tired at the end of the day, tapping all of the stuff. So and also, I think, I know, the focus on monetization, which is the big issue, which made sense in showing partners. You know, obviously, the next phase of 5G is going to be more about industries, the partners, than the operators themselves doing it, right? So that kind of made perfect sense.

Prakash Sangam

So, okay, let’s get started with the first thing they mentioned, programmability of networks, and now for critical enterprise needs and system needs and so on. So they had kind of two good examples mentioned there as well, right? Network slicing and energy saving. And I also liked how they showed about programmability is not, you know, only at one part of the network, it’s throughout and highly granular, right?

So going from millisecond response time in terms of turning on and off your PAs in the radios to term days and weeks, months kind of granularity at the higher level for network configuration and so on. So that I think was interesting. It’s not just talking about it, but also showing how it is done.

And one strange thing was that, you know, we all know Ericsson was a sceptic of OpenRAN in the beginning. Now with all this programmability that they’re talking about, which is riding fully on this openness and OpenRAN architecture. So they have become kind of the biggest proponents, right?
Especially after getting the AT&T contract.

Leonard Lee

Yeah. I mean, I think OpenRAN is there as a tier of interoperability that obviously all players outside of China need to contend with and deal with, right? I mean, I’ve said this for a long time. This is an operator’s choice, right? You can’t push open on people. Operators are going to make this choice individually or maybe as a coalition.

And there’s a lot of attempts to make this a regulatory requirement or embedded into standards and stuff like that. But at the end of the day, one of the challenges we see with Open RAN and O-RAN adoption is simply because operators just don’t want to. Or vendors like Ericsson and others are able to bring to market systems that are not as open, right?

The purpose built that are more performant and scale better. Now, I think what we still need to be very conscious of is that you can have programmability without any of the stuff being open, right?
This is really more of that conversation around having a virtualized base and then evolving that into cloud native and then having a cloud native architecture modularization of software around key elements and moving toward like sort of a service based architecture, right?

And then, you know, that all is enabling programmability of the infrastructure in different levels, evolutionary levels, right? And so to me, I thought the programmability stuff was less about open. It was more about how do you continue to modernize your network and fusing, you know, cloud native capabilities and architectures and practices and principles on top of a virtualized RAN base, if you will.

Prakash Sangam

Yeah, I agree. So it’s, I think, the combination of virtualizing the network, which is a basic, you know, and then if you want to really scale out and bring in all different kinds of players into the ecosystem, then you have to have open interfaces so that the guys who are developing, you know, application services and architectures top of that network have kind of portability between ecosystems, right?

Not just, you know, purpose built. So, yeah, yeah, I mean, Open RAN is not just for multiple radios sake, but it also for higher, higher levels of openness.

Leonard Lee

Yeah, yeah, yeah, exactly. And I just see it as just a fortunate thing for the Open RAN community, right? Because it’s, I mean, one of the things that we’re seeing are these like sort of, what do you call it, single vendor Open RAN deployments or, you know, pilots and implementations. I mean, this is where the general interest, I think, or the prevailing interest amongst operators lie.

That just simply becomes a benefit for the Open RAN community. It just at least opens that interoperability door so that, like you’re saying, these partners can participate. Depending on the operator’s decision, right? This is not a Ericsson decision. This is what the operator wants, whether it’s AT&T or who have you.

And so it’s interesting how Ericsson is playing kind of like this catalytic role, even though they’ve been kind of perceived as being sort of the wolf coming into the hen house. You know what I’m saying?

Prakash Sangam

Yeah. So I think the discussion has kind of moved on from mixing and matching radios and the bass band to more now on higher layers, right? In terms of bringing cloud native architecture and then bringing different vendors, especially the software and application and services vendors on that network. Where I think, you know, they have to work, the operators and vendors have to work openly with the outside ecosystem, the software developer ecosystem to really monetize this solid and excellent network they have built, right? So I think, and rightly so, all of the discussions during the day was on that, right?

Leonard Lee

So, yeah.

Prakash Sangam

And then, moving on to the next focus area was on APIs. So as you mentioned, there is GSMA working across the board to come up with common APIs for operators to employ, so that they have large economies of scale. And then Ericsson on its own signed a joint venture with tons of operators to take to the next level and accelerate the API adoption. So that I think was a key discussion and theme throughout the day, which made perfect sense.

Leonard Lee

So, yeah, and it is a big deal for Ericsson. But then I think also for the broader participants, the 12 operators as well as Google, GCP, who is one of the partners, right? The way that you gotta look at it is that the CSPs or the operators are kind of like the suppliers, right? And they’re providing whatever network capabilities and exposing those network capabilities to the APIs.

And so this whole GMP new co-venture is about creating an aggregation service, like a sort of an exchange, right? Or that will expose these network capabilities through APIs, right? Everyone calls them network APIs. And like you said, they’re being specified by the Camara project. It’s a Linux foundation project under the GSMA sort of purview of the Open Gateway Initiative, right?

And that was announced actually early this year. So a lot of the stuff is moving really quickly.
But I think the adoption, that’s something that we still have to see because I think there’s a general lack of awareness amongst developer communities about what a network API is, right?

Yeah.

And one of the benefits I think that Ericsson is anticipating with the Camara standardizer, if you want to call them open network APIs, is that getting scalability through that. Because one of the things that they really need to do, and this is something that they’ve stated, is keep the transaction costs low, right?

I mean, they need to be able to help operators expose these network capabilities and then have them made available and consumable by quote unquote developers at a low cost, right? 
And at least with economic scalability, actually, let me take that back, it’s not low cost, but with economic scalability, and that’s where those open standards or that open API set is going to be a huge utility, especially if you look at or consider the fact that eventually they’re going to probably be many more than 12 operators that are part of this network, but then you’re going to have to service probably tens of thousands, if not millions of developers, right?

Prakash Sangam

Yeah, absolutely. And of course, it’s very early days, but things are moving fast for sure.

I had the same question, but you asked during the event in terms of monetization of these APIs, especially for operators, right? I mean, I asked this question throughout the couple of times that all of this from a technical standpoint looks good, you know, large ecosystem play here, and there are opportunity to scale up.

But how is the whole ecosystem making sure operators which are at the foundation here are well compensated for all that they are enabling at the top? I mean, you know, 3G and 4G where operators built the network, but most of the value was taken away by the OTT players.

So my question was more on how the ecosystem is making sure that is not repeated and operator interest as well as everybody in the value chains interest are served and then the value is distributed proportionately and fairly among all of the participants, right? Not concentrated at one place, right? So, I mean, there are hooks and ways to do it, but I don’t see there is a strong blueprint, if you will, on how to make that happen. I think everybody is saying, let’s see how it goes and so on, right? So what are your views on that?

Leonard Lee

And just to be clear, I think that that was your question. You asked that question, right?

Prakash Sangam

Yeah. I think you asked about how operators will monetize APIs.

Leonard Lee

Well, yeah. And I think the monetization question for the operators is pretty simple. It’s you’re going to have access to a certain either net new market that you didn’t have before. And or you will be able to participate in, you know, existing markets where you’re going to have access to a larger scale of a market, right?

Just simply because this is like an aggregated pool of opportunities is the way that I think a lot of operators are going to be looking at this. So monetization, there’s monetization opportunities there, but the big question is on the demand side and then like what you’re saying, the how big is this pie?
And, you know, at what level is the bulk of monetization and revenue and profit going to eventually fall?

And, you know, I’ve written about this already several times and I’m fierce, you know, Dan Jones, he gets me involved in all these API discussions. And one of the things that I’ve cited is that, you know, one of the key layers, if you really look at this as like sort of a stack of, you know, a value stack, one of the key layers that needs to start to form or develop is sort of this what I call the network application layer, right?Because you have these APIs, they’re really probably going to be very transactional and, you know, sort of narrow in their functions, right? So if you go and you look at the Kamara API descriptions, they have like a portfolio of like 30 or so APIs.

These things tend to be pretty narrow in function. And where you’re probably going to see a lot of value kind of pool is at this network application level, where it’s either, let’s say, a standalone application of some sort. So let’s say that it may be a… might be, let’s say…Authentication, for example, right?

Prakash Sangam

Yeah, fraud prevention.

Leonard Lee

But it’s a suite of functions that tap into several, what do you call it? Network APIs. But that application stitches things together in a more value adding function, or even application, that may or may not plug into a business application. So that’s what we’re seeing with what you’re saying, authentication.

You have these new, more efficient, effective ways of authenticating applications, and they don’t need to be mobile, by the way, that leverage information and security capabilities of the network, right? Which is actually a really, really interesting paradigm that I think we might not be, a lot of folks might not be tuned into. So when you start thinking about that network application layer, that’s probably where a lot of that value adds going to happen.

But then I guess the hope is that that will drive demand for API calls that will then expand the revenue opportunities for operators. And these will probably eventually become like network evolve into net new and differentiated services. So I think that’s that dynamic that Ericsson and the new co-partners that they’re hoping for or anticipating.

Prakash Sangam

Yeah, and I think the question on what are the, monetization opportunities for operators, obviously one is for the access to the APIs. I did not get very clear answers from the operators on the panel on how much they are thought about it, which I think is not concerning, but I think worth noting maybe they should think a little bit more beforehand on how they want to monetize it, rather than just, you know, we develop and we’ll see how the monetization is later, right?

So, I hope they don’t see access is obviously one of the main things where they’ll make money on, and they have to make sure that they price it as such, that will encourage more usage than discourage usage. The pricing should be that, you know, cost effective to access it rather than make it, you know, high-end, very exclusive, right?

It should be, in my view, more on a revenue share basis, more access and they come up with more progressive and revenue share kind of a model than just fixed or something like that.

Leonard Lee

Yeah, and I think the key is, you know, pricing for value, and I think it’s going to be different for certain, let’s say, offerings that might compete with existing paradigms of, let’s say, you know, authentication security or network resource discovery. I don’t know.

That’s why I mentioned this whole network application layer, because that’s probably where the end user or the developer is actually going to interface with the network, right? And then underneath, that engagement is going to then translate into a whole, let’s say, collection of API calls and how that gets monetized, cleared, and settled down to the operator.

That, I think, is definitely not as visible at the moment. And I think are things that everyone needs to figure out, not only Ericsson, but everybody. And so I think one of the things that Ericsson definitely emphasized as well as the partners, was that this is going to be a journey. This is the starting point, right?

This is not like, okay, in 2025, first half, this thing is going to explode and it’s going to be the next best thing since sliced bread. This is going to take time. There’s a lot of design and execution still that needs to be done. I think that’s pretty apparent. The framework’s there, right? It’s just, how do you execute and position a GMP and the NewCo JV in a way that can serve as a catalyst and drive additional ecosystem value? That’s still a big question.

Prakash Sangam

It’s, I think, a tough one to answer. And my last point on this one is, for example, when the mobile broadband revolution started, oh, we will build and we’ll provide connectivity since everything writes on connectivity. We will have control over it and we can monetize in many ways, right?

That was the understanding, but the underlying connectivity was given. So value started building up on the top. So my point here is, shouldn’t be similar in this case, wherein, okay, the access to APIs is basic and is given. You can’t charge me more for this kind of API, what’s that kind of API. And the mobile broadband model where the commodity that you have access and you cannot differentiate shouldn’t be the case is my point.Okay.

Let’s move on to the next one.

So we had speakers from Meta, AWS, Google, and other partners, but they also an investor there, Greg Papadopoulos. Erik had a fireside chat with him, which I think was interesting.  I was encouraged to see that the venture community is really thinking that the APIs and the Edge AI itself will have a major role to play, especially when the AI workloads move towards inferencing and inferencing at the Edge, right? Than mostly cloud right now.

Leonard Lee

Yeah, and by Eric, you mean Eric Ekuden.

Prakash Sangam

Yes, Eric Ekuden of Ericsson, he had a fireside chat with Greg Papadopoulos.

Leonard Lee

Yeah, I thought it was really interesting. I mean, you know what? Greg mentioned distributed edge computing quite a bit, which I thought was really interesting. And he also brought up the term federated in the context of AI, quite a bit as well as portability. And so, you know, I think it’s a statement of a big problem to be solved, especially if you have enterprises and even, you know, cloud service providers, hyperscalers, that want to now propagate those workloads closer to where things are happening, right?

And where the services are consumed. And maybe entertaining, you know, new compute architectures that can not only deliver better experiences but also provide some better economics, you know, lower costs to serve, hence better margins, right? And it’s kind of like what I heard from Greg in terms of the motivation of looking at, you know, like, distributed Edge AI, computing and the network APIs. I think it’s just really early stages at the moment, based on, you know, sort of how Greg was talking about stuff.

But it was interesting to see how him articulate what he thought the role of these new network capabilities that are exposed through APIs could have in enabling these new architectures, right? Especially if you overlay, let’s say, AI, you know, there’s all this talk about AI, RAN and stuff like that.

But I guess it begs the question, what is the role of the RAN and the network and these APIs in bringing about these federated and distributed computing models for AI? And I don’t know, I kind of thought they were talking about generative AI because AI is already out there. You know what I’m saying?

Prakash Sangam

Correct. Yeah, generative AI at the edge. And I think the point he mentioned, which I’ve written a lot about as well, is basically you have the compute on devices, which is already paid for by the users.
So incentivizing for the AI folks to push as much processing to these devices or things on the edge, for which they don’t have to pay the compute.

So reduce the compute need at the Cloud, which is very cost-effective for them. So I mean, it was more of a device player pitch, like Qualcomm, MediaTek and others. But it was interesting to see that VC, who very often talks to and invest in the app developer ecosystem being bullish about that.

Leonard Lee

Yeah. It was interesting to hear.

Prakash Sangam

So the next one on our app ecosystem, there was a nice panel there as well on how app ecosystem is developing, what are the things in the use cases that they’re using primarily right now for OSS and BSS, and portability between different vendors and so on.
So I think our app is still at a very early stage. But that’s where I see, of course, APIs is one area where app developer ecosystem will have a major role in terms of coming into the 5G ecosystem benefiting from it. Even going back to this open RAN idea that even small companies can enter into telecom ecosystem.

I don’t know in the radio and the RAN space that is happening or not, but our apps is place where smaller nimble app developers can come in, provide very sophisticated applications and services, and kind of enter the large telecom ecosystem, right? Which was very hard for them till now.
Yeah, and I thought the talk about SMO was really interesting.

Leonard Lee

I don’t necessarily, yeah. So our apps themselves are probably pretty early in the adoption phase. Just simply because there are a lot of ricks out there.There’s folks that are dabbling with it and deploying SMOs and service management orchestration platforms and such. But I think this is an important entry point for a lot of companies that are looking at undergoing an open ran, let’s say open ran-ish transformation or let’s say revamping of their or evolution of their network infrastructure.

And yeah, it was interesting to see them bring this up as being an important tool and a way of rationalizing your portfolio of SON tooling and capabilities, right? I mean, and I think that, you know, I’ve written about this quite extensively and for a long time, that this is that integration hub, you know, like whenever you’re doing a transformation, one of the things that you look for are those integration hubs and middleware that allow you to have that kind of interoperability, flexible interoperability without necessarily having to have this perfect standardization of interfaces, right?

You have a tool that allows you to connect a system on one side of your network that might be of a certain technology, might be purpose built with something that is more open or standard space, you know, it gives you that flexibility. And so, you know, it’s interesting to see them bring that to the forefront.

And, you know, I think if they continue to gain some traction among operators who see the value of SMO and RIC as a target toolset, then you’ll see our apps start to flourish. And so, we’ll see, right? But we’re talking about rApps. A couple of years ago, everyone was getting excited about XApps, right? That didn’t quite fly, right?

Prakash Sangam

Definitely, our apps has more legs than XApps, right? So, and you’re right, when we started asking questions to people about SMO. People were still figuring out what it is, why do you need it and so on. But now there is realization that SMO is key and there are multiple players and infrastructure players themselves, third party players trying to get to that space. That will become really interesting. And that kind of becomes kind of a gateway into the network, right? That will be very strategic.

Leonard Lee

Exactly.

Prakash Sangam

Whether the traditional Telco players will play a key role there or third party software companies who have much bigger expertise there will rule that market. It will be interesting to see. SMO space is one hell of an interesting thing to keep looking at.

Leonard Lee

Yeah. Well, AT&T is definitely moving forward with it. So it’s a key enabler for their network trend.

Prakash Sangam

All right. So, and our last thing, connected PC enterprise pilot. So that was an interesting idea.
So basically Ericsson is doing a pilot wherein you replace all connectivity in enterprises with connected PCs. So, you take out your Wi-Fi and wire network in your offices and replace that with PCs which have cellular connected all the time.

So I think it’s an on-going pilot, but they presented interesting results wherein since you don’t have to maintain an IT network of your own, all of that cost is gone for you as an enterprise. And only thing we have is basically paying for a connectivity module in your laptop and the connectivity service from an operator.

So the pilot they did, they seem to show that it is economically more beneficial to just rely on the cellular network for all connectivity, not just when you’re traveling or something like that, but even for your in-office connectivity. So we’ll see how that goes, but that’s an interesting idea.

On the face of it, Wi-Fi looks much cost-effective and cheaper compared to cellular.
But when you look at all the equipment you need to maintain it, people needed to maintain it, upgrade costs and so on, Ericsson claims will end up becoming higher than having a cellular connectivity.

Leonard Lee

Yeah. It boils down to this on-premise infrastructure and having to maintain that. Looking at not just Wi-Fi, but holistically of managing and hosting your own network locally versus what they call infrastructure-wide. Hey, look, the operator already has infrastructure up, they have spectrum, you have the devices.

Rather than doing any of this build out that you’d have to do to host your network and manage your network on-premise, why don’t we just use the operator’s network and you just piggyback on top of it.We’ll provide you with tools and the cloud-based, if you will, administration platforms to manage your network without having to deploy anything on-site. You’re going to be just connecting to cell towers and or that local cell tower that is going to service your particular facility or office.And so it’s a bold vision.

I do have questions on what sort of scenarios and what kind of companies would this value proposition appeal to? Who would actually entertain this? Because this is kind of a mindset shift, right?

Prakash Sangam

Yeah, it’s making a cellular primary connectivity vehicle, right? So, but I think it’s an interesting idea.
Of course, it is self l-serving for the cellular community. Yeah, but it is an interesting idea, especially capacity on these networks is increasing so much. For many use cases where users are not heavy broadband users, like office connectivity,  if you’re an average user, it’s not that crazy, right?

We’ll see how the results come out. And of course, the issues of you have to have perfect coverage everywhere. What happens if there is no coverage at the corner of office?

Leonard Lee

Yeah. I mean, you know, I don’t know. You know, some of the things that popped into mind were like pop-up stores. That might be perfect for this. Also, you know, like a WeWork-ish kind of office for a buyer type of environment. For some of those companies that are a little bit more mobile and maybe scale, scale up and down, you know, as part of their business. That’s where this type may appeal, but it’s novel.

Prakash Sangam

Although they were suggesting I talk to a few people who are in the pilot on the sidelines, their target is not just special cases. The target is any average enterprise like office building, not a shop or stuff like that. And this is basically you have a laptop with 5G connectivity and no Wi-Fi in your office. We’ll see how that goes.

All right, so I think we talked enough, right? A lot of interesting ideas. So do you  want to give a plug on where people can find your research and your insights?

Leonard Lee

Yeah, sure. Just go to www.nextcurve.com. I have a media center there where you can find access to all my media content, as well as a research portal where you can access my publicly available research.

Prakash Sangam

Okay. All right. that was a great discussion, Leonard. Thank you for sharing your insights, and thanks again for coming on our show. We’ll see you again soon.

Leonard Lee

Thanks for having me.

Prakash Sangam

Sure. So folks, that’s all we have for now. I hope you found this discussion informative and useful.
If so, please hit like and subscribe to the podcast on whatever platform you are listening this on.
I’ll be back soon with another episode, putting light on another interesting tech subject.

Bye-bye for now.

 If you want to read more articles like this and get an up-to-date analysis of the latest mobile and tech industry news, sign-up for our monthly newsletter at TantraAnalyst.com/Newsletter, or Read our Tantra Analyst Insights Articles.