Semiconductor innovation: The impact on telecommunication
The telecommunications sector is increasingly becoming characterized by fixed operator CapEx and increasing user demands. It’s a dynamic that’s not sustainable, and something needs to give.
In this latest Fierce Wireless digital interview, we’re joined by two individuals with a unique perspective. Speaking with Joel Brand and Stephen Adolph of multibillion-dollar technology and semiconductor giant Marvell, we deep dive into the criticality of chip innovation in the telecommunications sector.
Setting the scene, Joel both highlights the hard time that carriers have had in realizing the value since transitioning from 4G to 5G, and points to the growing shift towards exciting services that will demand more wireless connectivity.
“This short-term reality is opening the door for new players like the hyperscalers, like the satellite providers, to bring some new solutions and new possibilities,” he explains.
Attention then turns to the role of fixed wireless and private 5G networks as the industry transitions, alongside other key technical trends such as advances in silicon photonics and pluggable coherent optical transceivers.
“As we invest in next generation versions of this technology, we’re getting more and more capability out of smaller and smaller devices,” Stephen explains.
For more on both the perspective of operators and the way in which technological advances are helping to address their pain points, check out the full interview.
Alejandro Pinero:
All right. Welcome everyone to another Fierce Wireless Digital Interview. As always, my name is Alejandro Pinero, your host and a team member here at Fierce Wireless, and it’s my pleasure today to be joined by not one, but two industry experts. Joel Brand, Senior Director for the processor business at Marvell, as well as Stephen Adolph, AVP, CDSP, product line management in Marvell. Welcome, let’s start at the top, and I was hoping that you could give us a brief update on what are some of the key issues today when we think about chip sets in telecoms, whether general purpose, special purpose processors. Marvell, of course, has great experience in this topic, so let’s give a bit of background.
Joel Brand:
Sure. The reality in the wireless market is that operators for the last few decades have been operating with fixed CapEx, and we as consumer demand a lot more from their network. We want faster speeds, wider bandwidth, we’re using more sophisticated equipment and services, but that needs to be somehow addressed, right? Somehow the equipment needs to provide all these services while keeping the cost down and while keeping the power consumption down. So these are the key elements of designing chips for this industry, and that’s why Marvell is so focused on putting a lot of these capabilities, what we call accelerators, domain specific accelerators, in hardware. And the focus again, is to be able to service the demand while keeping the cost down and the power consumption down.
Alejandro Pinero:
Absolutely. As always, there’s a lot going on in our industry, and I’ll certainly ask you, Joel, in a little bit about some of those applications that you were mentioning there because I’d love to get your thoughts. But before we get into the detail, I did want to ask, from the carrier’s perspective, keeping in mind these requirements and the greater needs and expectations from customers and what that means for the networks that they’re running. What does 2024 look like for them? And perhaps even looking ahead towards 2025.
Joel Brand:
So in general, again, this market is characterized by fixed CapEx from an operator perspective and increasing demand from a user perspective. And my crystal ball is no better than anyone else. I don’t know when the next iPhone will come from, whether it’ll be a metaverse, whether it’ll be some hologram phone that requires a lot more. I don’t know where it’s going to come from, but it will come from, because that’s the nature of the world we live in. And that will increase the demand for wireless connectivity, for lower latency. We will see that coming. Whether it’ll come in ’24 or ’25 or a few years into the future, I don’t know, but it will come.
In the very short term we are experiencing, or the operators are experiencing, the fact that they have hard time realizing the value on the transition from 4G to 5G, and as a result, we are seeing in the market that their vendors are suffering in the market as a result of it. But again, if you look at the longer term trend, it’s toward more exciting services that will demand more wireless connectivity. I think this short term reality is opening the door for new players like the hyperscalers, like the satellite providers, to bring some new solutions and new possibilities, how this market is going to evolve in the future.
Alejandro Pinero:
Sure. And how then should these carriers think about how they’re building their networks? I mean, you’ve mentioned their already open networks, you’ve mentioned satellites, you’ve mentioned the hyperscalers. Is there a rethink about how to go about the very foundation of how we put cellular networks out there and working?
Joel Brand:
Absolutely. Absolutely. And you see that some of the leading companies in this industry, if you think of companies like SpaceX, if you think of companies like Amazon, that are launching satellite wireless services. If you look at the announcements that were made last MWC by all the major players in terms of partnerships with the hyperscalers, with server vendors. Absolutely.
Alejandro Pinero:
And Joel, before we wrap up on thinking about these new applications, I wanted to throw the idea around fixed wireless access. Something that we’ve been covering a lot here at Fierce and seems to have had a pretty breakout year, I would say, in 2023 and perhaps into 2024 in terms of its adoption and some of the results coming out. Private 5G, something that we’ve been hearing about a lot as well. These are things that we’ve been hearing about for a while. Is this really the time when we’re going to see some serious adoption and use?
Joel Brand:
Again, my crystal ball is not any better than anyone else, but we see movement. We see, again, because operators and vendors need to find new sources of revenue streams, we absolutely see the enterprise market. You see a lot of players that are trying to find solutions. And for example, Marvell is working with Dell on wireless solutions. And one of the target markets is the enterprise market, given Dell presence in this market. It’s a matter of focus and investment. There are wireless enterprise private network solutions out there. Just like the industry invested a lot in building solutions that are tailored for the consumer market, we can’t just assume that the same thing will work for the enterprise market. And as an industry, we need to take that head on and adjust the offering to better match the needs of enterprises.
Alejandro Pinero:
Excellent. So let me shift over here to Steve because we’ve heard a lot about these applications. We’ve heard about everything that’s going on in terms of wireless services, but there needs to be a backbone there to deliver on all this. So Steve, your group over there at Marvell focuses more on the optical side. What does that mean? What are you doing there as a company? Where are you researching and investing?
Stephen Adolph:
Sure, thanks. Well, at Marvell, our focus is to enable, what I’ll say is the development of best in class optical plugable modules. And a plugable module is a little device that’s small enough to fit in the palm of your hand. And generally people use those to interconnect networking products. Whether those are connecting inside a building like in a data center or connecting between buildings like in data center applications, or in this case of telecom infrastructure, they’re used to connect metro equipment together in the access, also in regional long haul networks. And you even find them in such applications as submarine networking where people are trying to interconnect over sub-sea fiber. So that’s Marvell’s focus, is those applications. What we develop fundamentally are silicon digital signal processors that are optimized for the use case. So there’s quite a variety of signal processors.
Generally we think of them as either, what we call PAM4 signaling or they’re based on coherent technology. And my focus is the coherent technology.
Alejandro Pinero:
Great. And I think it’s fair to say that this is an area of research and investment that’s been going through a lot of change and a lot of progress in terms of those technical capabilities. Can you maybe talk a little bit about what some of those key trends are?
Stephen Adolph:
What’s been going on in the past five to six years is that coherent technology is actually, due to advances in both the silicon photonics as well as semiconductor technology, the implementation of the ability to build a coherent optical transceiver has actually shrunk in size and shrunk in power and shrunk in size to the point where you can actually create a plugable version of these coherent transceivers.
Historically, these coherent transceivers have been pretty large devices and really implemented on almost a circuit pack level, so they’re quite large and you wouldn’t think of them as plugable. So that’s an important trend, this miniaturization of coherent technologies is what’s really allowed the development of plugable coherent products. And we see a tremendous amount of growth in that area because it provides a couple of significant benefits to network operators, whether they’re in the data center or in the public infrastructure. What you can do with these plugs is you can reduce your system cost significantly because these plugable modules, they do a couple of things. They allow you to, in some cases cost reduce the footprint of the optical equipment, and in some cases, they allow you to simplify your network, allow you to eliminate other platforms from the network.
Alejandro Pinero:
And on that point, have you already seen an impact then in terms of the amount that’s being spent by carriers on optical and plugables?
Stephen Adolph:
Well, certainly the adoption of coherent plugables has been in full swing in a couple of segments of the market. I think increasingly the demand is taking off. One of the big innovations or crystallizing events in this industry was a few years ago, the introduction of 400 gig in plugable form factors really started what I’ll call the revolution of IP over DWDM. I’m sure that’s not a new term for any of us on this call here, but what it sort of implies is the ability for a network operator to move the high performance optics directly into the router platforms and get a much more cost-efficient metro and access deployment. So that started off with the introduction of 400ZR and ZR+ plugables. We’re now on the cusp of transitioning to 800 gig technology. Marvell recently announced the availability of the industry’s first 800 gig coherent plugable DSP called Orion.
That’s enabling Marvell and Marvell’s customers to bring this new generation of 800 gig plugs into the market. And those really promise to do two things. They double the capacity relative to the prior generation update under gig, but they also enable very long distance 400 gig transmission. So many of our customers are looking at this as a double benefit. They get to solve their capacity challenges in metro with 800 gig, and they get to solve a much lower cost solution for a 400 gig transmission in regional and long haul. So those are some big advancements in the industry, and we expect this trend will continue. So as technologies mature, as we invest in next generation versions of this technology, we’re getting more and more capability out of smaller and smaller devices. So we see this trend continuing and we think that the focus of innovation and coherent optical, we’ll be grounded in enabling these next generation plugable coherent modules.
Alejandro Pinero:
Excellent. Well, I’m afraid we’re just about out of time here, but I think this has been a fascinating conversation and really an end to end conversation about all these upgrades and everything that needs to go into to make sure that we’re still delivering as a telco industry, but not spending too much on it or more than necessary. That’s a key takeaway. Well, Joel, Steve, thank you so much for joining me here on this interview. I have certainly learned a lot and I’m sure our Fierce Wireless watchers and readers have as well.
Joel Brand:
Thank you. It was a pleasure.
Stephen Adolph:
Thank you.
Alejandro Pinero:
Great. And thank you also for watching this video interview here on Fierce Wireless. We’ll be back with more content your way very soon. Until then, bye bye.