When I kicked off the Telco in 20 podcast almost two years ago, I wanted to claim a little corner of the podcast universe to explore how the telco industry is moving to, and using, the public cloud. My idea was to talk with hyperscalers, vendors, and even telco execs themselves, about how they’re redefining the way telco services are deployed.
I named the podcast “Telco in 20” because I was going to talk about the next 20 years in telco, AND I wanted to make each episode about 20 minutes long (so people could listen easily in the car or while on a quick walk!).
With the latest episode, the podcast is reaching new heights! I had the opportunity to sit down with DISH Wireless Executive Vice President and Chief Network Officer Marc Rouanne to talk to him about DISH’s new 5G network. As astute readers will recall, Marc, along with DISH Chairman, Charlie Ergen, announced a year ago that DISH is deploying a new network in the United States and building it on AWS, while spending only $10-billion dollars to do it.
Almost immediately everyone said, “DISH is crazy! It’ll never work!” But just like Elon Musk on his mission to Mars, Charlie, Marc, and their team weren’t afraid to be telco pioneers. DISH is embarking on the journey to make its vision a reality and attempting to set a new standard for how networks operate. It’s apt that the DISH headquarters is in the western United States, known for pioneers and gold rushes.
I hope you will WATCH this video of the really great conversation I had with Marc and hear us talking about true cloud innovation in telco. For this recording, we conducted an in-person interview at the DISH headquarters in Littleton, Colorado. You can catch the video on the new TelcoDR YouTube channel, but if you don’t have time to watch it, at least read the transcript of our conversation, which I’ve included below. Enjoy!
Transcript: Telco in 20 | Ep 44 – DISH: The Telco Pioneer
DR: [01:46] Marc Rouanne is Chief Network Officer and EVP at DISH Wireless. Hi, Marc. Welcome to Telco in 20.
Marc: [01:52] Hello.
DR: [01:53] I’m so excited to talk to you. I mean, when I put together the Telco in 20 podcast, which focuses on the next 20 years, and what you’re doing at DISH is exactly what I was thinking. I can’t think of a better guest for our podcast. So to start, I want to focus on your vision.
[02:08] DISH’s approach with its 5G network and its work with AWS is going to change how networks are deployed in the future. You’re taking new enabling technology, in this case, open RAN and AWS, and you’re creating a new telco business model that makes what was formerly very complicated and super expensive, now easier, faster and less expensive. If successful, it will set the new standard in the industry on how to roll out networks.
[02:35] So, do you see DISH’s network strategy as a totally innovative tech project that in 10 years, we’ll be saying, “Holy cow! DISH set the standard for the way networks are built,” because that’s how I see this project.
Marc: [02:49] Yeah. Actually, it goes well beyond DISH. I think we are lucky to be consuming technology now that is available to us and it’s completely new in the telco environment. You mention it’s the cloud, it’s new types of software. You have to see that all the cloud guys have been investing very successfully for 15 years, and for some reason, the telco never really tapped into that. Now, all of a sudden we can, and it’s almost unfair because, all of a sudden, we inherit 15 years of massive investment and capabilities.
[03:18] The other thing that is coming to us, where again we are lucky, 5G was designed by the standardization people to be cloud-native. So if you look deep into the 5G standards, you see everything for microservices, everything for data-centric networks, whereas 4G, 3G were never designed for the cloud. So it comes at the right time and we’re surfing that. So yeah, it’s brand new and very different.
DR: [03:43] I think you touched on something there that I think is very important. There hasn’t been another industry that’s invested as much CapEx as the hyperscalers. Usually, it’s always been telco that’s been the big CapEx people. Now you’re combining these two things together and you get to ride their investment as much as they’re riding yours so I think that’s another great insight.
Marc: [04:04] Yeah. The big investment they have made as an ecosystem is around silicon and stacks, which the telco never really did, right?
DR: [04:12] Yep.
Marc: [04:12] So we were a bit starving in the telco from a lack of very advanced silicon and software stacks. All of a sudden, now this is coming to us so yeah, it’s a new game.
DR: [04:22] Yeah. So there, you’re talking about Graviton2 and 3.
Marc: [04:24] Yeah, Graviton, but also Nvidia, Broadcom, Intel, Marvel. They’re all investing in ways that we’ve never seen before and this is pretty cool.
DR: [04:32] Yeah, it’s super awesome. So DISH is the first telco to really use AWS in this novel way. If you could wave a magic wand and change one thing of the AWS platform, what would it be?
Marc: [04:46] I think I’ve said it a few times, we looked at all the cloud and when we started with AWS, they were not telco-ready, especially when it comes to networking, SLAs, all the different ways of transporting telco. So we’ve done a lot of work with them and they are upgrading that platform for us to be telco-ready and telco-grade. But suddenly, if we had come when they were already telco-grade, that would’ve been easier for us.
DR: [05:11] Yeah, yeah. So you’re forging new ground and it’s a learning path. I think with that, there’s going to be setbacks and delays, things on paper look easier than when you actually do it. But I think when you look at the long arc of time, you guys will work through this and get to that telco-grade state.
Marc: [05:29] Yeah. But you know it’s true, there are a lot of things to do. But on the other hand, I’ve been a vendor for decades and my technical work here is much easier than it was before. The reason is that I was carrying so much luggage and complexity for putting old stuff into the current deployments. It was a nightmare, and I don’t have that anymore. So yes, there are new grounds, but everything we do is much simpler.
DR: [05:53] Right. Well, software is much simpler. It moves much more quickly than hardware and so that’s my next question.
[05:59] I’m a software girl. We’ve seen a lot of vendors on the software side in telco claim or say that they’re cloud-native. I mean, you’re actually trying to deploy it on the public cloud and so as you put together the software, you need to deliver this kind of new network. What do you need from the software players to help you complete your vision?
Marc: [06:19] Again, because we were starting new, we put very strong requirements on being cloud-native and we told everyone before we launch, which is now, you have to be cloud-native, and I was a bit skeptical. But they did it, and they did it faster than I expected.
[06:34] Now we may see over time that we still have some bugs to correct our learnings, but I must say I have been impressed by all the software vendors. Because again, the cloud market is mature so for them, it’s easy to have engineers that understand cloud-native. It’s easy to understand what is a stateless machine, what are the CNI they need to use. All this is actually at scale so they have done it pretty well. I think most of them understood that they had to rewrite their software, start from scratch, not try to carry over the old stuff. That was a bit of debate we’re having at the beginning—are they going to bring us the old stuff and make it look like cloud-native? But I would say most of them didn’t do that so we are fresh.
DR: [07:14] Right. Well, that’s really hard for a software vendor because you have your legacy estate and your customers where they are, which in telco is mostly on-prem. Then also developing the software that you need that’s cloud-ready, not just, like you said, cloud-ported. That’s a lot of investment. So for the smaller vendors, they might struggle. Bigger vendors might be able to have the R&D capacity to pull that off so that’s interesting as you deal with that.
[07:39] That kind of brings me to cost. What do you see as the cost savings of building your network and IT systems on an AWS stack versus the old school approach of VMware and Oracle? Do you expect these truly cloud-native solutions to be 10%, 50%, 90% of the cost that you saw previously?
Marc: [07:58] I’ll talk about price later. But, first, the cost of running a software that is cloud-native, of upgrading a software, of maintaining it is much cheaper than the cost of having legacy.
[08:10] Now this is cost. Of course, there is a tension of price. Do you have people like AWS that have a price that is too high, or that it’s not what you want, or somebody else as an ISV? So there is a tension of procurement and customer. But if you forget that, the baseline is much cheaper. It’s cheaper, not only in terms of the cost of what we buy, like IT or computers, but also the OPEX that we have to put in is very different.
[08:38] Then it’s a play of the ecosystem. Who has the negotiating power and so forth? So far, we’ve been in a good place because as a trendsetter, everybody is trying to give us-
DR: [08:46] Must be a part of it.
Marc: [08:47] Yeah. They want to be part of it. They have given us good deals. But, of course, we’ll see how the market evolves. There may be places where, for example, for very massive compute, we have to in-source back into our old private capabilities when the scale is there so we are open to that as well. I mean, there is a tension that is going to be created between price and cost.
DR: [09:05] Right. I think that the price should be coming dramatically down. Again, speaking with my CEO of Totogi hat on, just on charging versus on-prem, not having to pre-provision the capacity and the failover-
Marc: [09:17] Absolutely.
DR: [09:18] … and being able to do that dynamically and elastically with AWS is huge. It just changes the economics for people. Software that previously was only available to the world’s largest telcos, now you can go a lot further down market to tier twos and tier threes and can now afford it. I think that’s going to change the landscape in telco. I’m super excited about that.
Marc: [09:38] It’s also the investment cycle. You can invest later when you have revenues and when you have traffic. And you make fewer mistakes because if you invest too early, you are on the very early generation, the bleeding edge of technology or silicon. Whereas we can just wait and take it when it’s ready.
DR: [09:54] Yeah, when it’s ready. Yep.
[09:56] Now switching a little bit to the enterprise market. AWS has become the dominant public cloud vendor and US IT companies inside those groups, but still we see telcos building their own private clouds with their proprietary APIs. Are you excited to compete with open APIs in an open stack versus your competitors that are still walking in with a private cloud and proprietary APIs?
Marc: [10:19] Yeah. I have a very simple principle. I always want to publish our APIs and capabilities and the more of our competitors that adopt it, the bigger the mass market, the better the economics for everyone. So I want everybody to join open APIs. I want everybody to publish and I will continue to publish everything we do because, again, you want to belong to the mass market.
DR: [10:39] Yeah. Well, I think the interesting thing here is enterprises. Your customers have already selected AWS. It has 50% market share and probably in the United States it’s pretty strong. So when those teams are comfortable with AWS and already using it, I think since you’re coming to the table with open APIs and AWS, it just makes that work so much more easily.
Marc: [11:01] Absolutely.
DR: [11:02] They don’t have to learn everyone’s different network and your APIs are published. Again, those IT people, they just want to use something that’s easy to use. Finally, you’re VP of enterprise sales. I’m like, “Give me AWS twice on Sunday. This is amazing.”
Marc: [11:16] I think you mentioned it. It’s all about developers and the ease of consumption. So yeah, if there is a big developer ecosystem, we’re going to need to adopt everything out there.
DR: [11:25] I mean, they’re just going to eat that up. Published doc, they just want to start coding, they start using and-
Marc: [11:30] Absolutely.
DR: [11:31] … they don’t want the hassle.
[11:32] All right. Now switching a little bit to competition. The other disruptors we’re seeing in telco are the cable guys. Their advantage in the market is their cable plant. They are in-ground cables already. So when you think about your strength that you’re bringing to the table with AWS’s CapEx and infrastructure, is that helping you as much as the in-ground cables are helping the cable guys?
Marc: [11:53] We have a number of assets. First, we have the Spectrum, so the Spectrum allows you to have access to mobility and devices and data. It’s a game about data. You need to have access to data and then you’re smart.
[12:04] But yes, the fact that we can go through AWS, its ecosystem of developers, of millions of developers, but also the frame of all the data centers they have in the US, it is a big asset for us in terms of SLA, redundancy, reliability, scalability so we are leveraging that.
[12:20] I mean, additionally, we like to do things where we can differentiate, so SLAs, software, but we never duplicate what people are better at. So cloud providers are good at cloud. We consume that.
DR: [12:32] Well, I think a really good example is 1&1 in Germany is also building a new 5G network, but they’re building all their data centers. They plan to build at least four major data centers and then all these micro data centers. So I think what’s really interesting what you said, you get to ride on the coattails of the most dominant public cloud vendor and I think that’s a huge advantage for you guys.
Marc: [12:53] And going forward we are very successful, we have a lot of traffic, we will measure the need to in-source some of that, do our own computing.
DR: [13:00] Yep, balance it. Yeah.
Marc: [13:01] But right now there is no need for that. We’re much faster using the cloud the way it is.
DR: [13:04] Yeah. Oh, it’s great. It’s awesome.
[13:07] Now sort of pivoting and speaking of competition, I hear that you’re big time into sailing. I read that in your spare time, you’re working on AI technology to give insights to teams during races. So I have to ask, are you using the public cloud for that? Please say yes.
Marc: [13:22] Yeah, yeah, absolutely. Again, it’s all about data and once you have the right data, you sail much faster. So yeah, it’s the same story, same family.
DR: [13:31] Yeah. It’s so awesome. Well, Marc, this is such an amazing conversation. Thanks so much for coming on the podcast.
Marc: [13:37] Well, thank you.