In this episode of The 5G Factor, our series that focuses on all things 5G, the IoT, and the 5G ecosystem as a whole, we review the early market inroads and impact that Telco GenAI is delivering. The major takeaways consist of BT Group using Amazon Q’s coding companion, Amazon Code Whisperer, to provide GenAI coding assistance to its software engineers, Amazon and Microsoft combining Amdocs amAIz platform and Microsoft Azure OpenAI Service to bring verticalized GenAI to telcos, and the debut of the Qualcomm Snapdragon X80 5G Modem-RF system that integrates a dedicated 5G AI Processor and 5G-Advanced-ready architecture to achieve key breakthroughs across NB-NTN satellite communications, 6-antenna architecture for smartphones, 6X carrier aggregation, and AI-based mmWave range extension for FWA CPE.
Our analytical review spotlighted:
Amazon CodeWhisperer Shows BT Group Early Benefits. At MWC24, BT Group announced that it deployed Amazon Q’s coding companion (Amazon CodeWhisperer) to provide generative AI coding assistance to its software engineers. Specifically, Amazon CodeWhisperer generates code suggestions ranging from snippets to full functions in real time in the customer Integrated Development Environment (IDE). The solution is already providing 15-20 suggestions of code per active user per day for BT Group, with an acceptance rate of 37% by its software engineers who are using the platform. We see this as an encouraging example of the value, efficiency, and support generative AI can offer to a communication service provider (CSP) organization.
Amdocs and Microsoft Ready GenAI Specifically for CSPs. Amdocs expanded its strategic alliance with Microsoft to further bring verticalized GenAI capabilities to the telecommunications industry, combining Amdocs amAIz platform and Microsoft Azure OpenAI Service. The duo has established a dedicated lab and center of excellence to further productize GenAI capabilities with the goal of accelerating adoption of GenAI technologies for CSPs. The collaboration is designed to bolster business outcomes by combining deep expertise in the communications industry with GenAI technology and industry verticalization that can produce carrier-grade business outcomes. Amdocs has built Customer Engagement Platform, which includes multiple copilot products, including sales, order enrichment, care, proposals, and other key functions, like configure price quote (CPQ), which we view as integral to providing an interconnected GenAI-infused experience across the entire CSP organization.
Qualcomm Snapdragon X80 Ties Together AI and 5G Progress. During MWC24, Qualcomm unveiled its new Snapdragon X80 5G Modem-RF System. The offering features new AI innovations based on a dedicated tensor accelerator to improve data speeds, latency, quality of service (QoS), coverage, location accuracy, spectrum efficiency, power efficiency, and multi-antenna management. it integrates a dedicated 5G AI Processor and 5G-Advanced-ready architecture to achieve what we see as several important industry milestones, including the first 5G modem with fully integrated NB-NTN satellite communications support, 6-antenna architecture for smartphones, 6X carrier aggregation, and AI-based mmWave range extension for fixed wireless access CPE.
Watch The 5G Factor show here:
Listen to the audio here:
Disclosure: The Futurum Group is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of The Futurum Group as a whole.
Transcript:
Ron Westfall: Hello and welcome everyone to The 5G Factor. I’m Ron Westfall, Research Director here at The Futurum Group, and I’m joined here today by my distinguished colleague, Olivier Blanchard, our Research Director, focused on key areas such as devices and semiconductors, and that includes 5G, naturally, and today we’ll be focusing on the major 5G ecosystem developments that have caught our eye.
And specifically, we’re going to focus more on GenAI and its impact on the mobile ecosystem, including especially telcos and what can now really be characterized as the telco GenAI market segment. And so Olivier, welcome back to the 5G Factor and many thanks for joining today. How have you been bearing up between our episodes? And that includes. Well, your wedding and again, a hearty kudos on the nuptials.
Olivier Blanchard: Yeah there’s the little ring right there. So thank you. Yeah. I just got married.
Ron Westfall: Outstanding,
Olivier Blanchard: So, I’m getting back into the swing of things. Yeah, no, it’s, I took a, almost a week off. And that was nice. I came back with a cold. You can probably hear it in my voice. I don’t usually sound this sexy. But I’ll take it. It’s one of the cool things about getting a cold is I sound like this and that’s, I usually record my voicemail message when I have a cold.
Ron Westfall: Well, you could sound a bit more authoritative perhaps. And if it’s any consolation, something is going around. I just had to deal with one myself. That aside again, it’s just great to have that ring and perhaps yeah, we can work that into our narrative today, .
Olivier Blanchard: Sure.
Ron Westfall: The Ring of GenAI and what it means for the entire mobile ecosystem. And so with that, hey, let’s jump right in because as we know about Mobile World Congress 2024 was almost a month ago by now. And and as such, we can describe 2024 itself also is the year of AI with GenAI taking a starring role. I think that’s a pretty broad consensus, and the mobile ecosystem naturally is no exception. In fact, some folks are suggesting that Mobile World Congress should rebrand itself as AI World Congress. However, I would not hold my breath on that exactly developing anytime soon. But with that in mind, while AI and GenAI were certainly the dominant themes that were driving the conversations there, there is some concern and criticism out there that it’s already being overhyped.
And naturally, our conversation will address that. I don’t think it’s actually being overhyped simply because for one thing, it’s at the front end. It’s still only a year old. And so I think we’re going to really see a lot of, I think, creative ideas, partnerships, innovation coming to the forefront. And bear with it. As we sort through, why it’s important to the mobile ecosystem, let alone the entire tech sector. But I think part of the concern was that it was drawing attention away from, key 5G initiatives in areas such as three GPP releases, 16 through 18, and, the monetized dimension is still not there yet. But again we’re at the beginning of the journey. And so some of these things, I think, aren’t going to be ready yet. But also, I don’t think it’s really distracting from the important groundwork that’s going on in 5G itself.
In fact, one area that I think GenAI can have a near term impact on, Tangible impact is an area that I don’t think has gotten it due is GenAI coding assistance. And at the show, BT group announced that it deployed Amazon Q’s coding companion known as Amazon Code Whisperer to provide GenAI coding assistance to its software engineers. Now the solution is already providing 15 to 20 suggestions of code. Per active user per day for BT group. And with an acceptance rate of 37 percent reported by software engineers who are using the platform. I see this as an example already of where the value efficiency and support. port that generative AI can offer to the telcos specifically. Now it’s foreground. Amazon Q. It can be tailored to a business by connecting it to company data, information and systems made simple with more than 40 built in connectors. So in essence, AWS is coming with Amazon Q and saying, Hey, look, we’ve done some of the heavy lifting already.
This is meant to be set for your business needs and not, having it shoehorned into, okay, if you want to onboard, you have to, strictly comply with AWS, Amazon queue requirements. Now, as such, in addition to the coders, they can support business users, such as marketers, project program managers, sales reps, and quite simply, they can have tailored conversations that solve problems, general content. Plus, recommend and enable actions. So it’s not just for the coders, at least the Amazon Q part, but also for, key decision makers across the entire organization of BT Group is coming to mind now drilling down specifically. Amazon Code Whisperer generates suggestions ranging from snippets to full functions in real time in the customer integrated developer or development environment, and that’s also referred to as IDE.
Now, in this case, BT based on their coders comments and existing code, it can also support command line completions and natural language to bash translation in the command line. Also, I think it’s important to note that for my view, code whisper is making a difference because it can also flag or filter code suggestions that align with publicly available code using open source project repository information and license for review and attribution. So right away at the starting gate, we’re seeing where it’s having impact specifically with one major telco. And it’s not just, GenAI in terms of a conversational natural language capability for, say, customer support, which is also important. Naturally, we’ll touch on that. But also Olivier, from your perspective, what is, really important that’s coming out of in terms of how the telcos are taking advantage of GenAI capabilities, at the starting gate here.
Olivier Blanchard: Yeah. So we it’s funny, we started with the not Generative AI specifically, but AI and telcos in terms of managing power and optimizing connections with even just millimeter wave RF systems, right? So we saw the start of AI integration into telcos with these sort of like meat and potatoes, but extremely complicated engineering problems being solved by AI on chip. And obviously on the other end too, on the antenna side. And now we’ved just move forward into a problem that seems actually simple or simpler to solve just because it’s solved more by software, And on a bigger scale, and this generative AI thing essentially like it takes over tasks that are very naturally, very complex, very like expensive in terms of time and resources and shrinks them down.
And so this no code thing or low code I would say actually just prompting. Or switching from code to prompt, and I think this is like the trend that you’re alluding to that we’re seeing pretty much across the board with everything. And I’m reminded of a presentation at HP Amplify by NVIDIA CEO, which he’s repeated and expanded on at GTC last week when I was, on on my little short honeymoon, I missed a lot last week, apparently. And it’s basically this trend to go from coding to prompting. Like what’s the new thing? What you prompt is what you get. I think that’s a trend across the industries. And telcos are on the bleeding edge of this for obvious reasons, just for scale, if nothing else. I think that Amazon doing this is, a natural fit for their business problem their business model AWS anyway.
And you’re going to see that kind of spread across the industry. I think everybody’s doing this. We’re also seeing a big push by the chip manufacturers whether it’s Qualcomm, AMD, Intel and others and NVIDIA, obviously to open up their platforms to developers. And one easy way of doing that is by enabling these kind of like low code and no code implementations. So that anyone with, very little training can essentially just learn to prompt what they want to build and just tell the system what they want the outcome to be and tweak things with prompts rather than with code. And the AI takes over and does the work for them. And I think it’s a fascinating. It’s a great time to be alive, right? But it’s a fascinating sort of path to democratizing.
GenAI and coding where it’s, it just, it reduces barriers of entry so that people who don’t necessarily have the time, the resources, or even the capability to learn how to code can still create apps and create all sorts of cool products and and solutions for themselves or for their companies, or, to monetize using this ecosystem. I think it’s wonderful. It’s actually one of the cooler things about the GenAI revolution so far. Cause there, there are a lot of things that I worry about with GenAI, but it’s definitely for the most part, I agree with you, not overhyped. I think in some areas it might even be under hyped, especially the speed with which it’s evolving and the speed with which we’re not coming up with safeguards for some of this stuff. But I think that this low code, no code approach is is one of the cooler aspects of this GenAI age that we’re entering, whether we like it or not.
Ron Westfall: And those are excellent points, Olivier. In fact, that triggers two immediate points. Thoughts on my side and that is yes, AI ethics, responsible AI, and we’re going to see more of this being built into, organization, decision making, how the company is, literally organizing itself, but also, at the board of directors level and, certainly the CX level, this is the, how the organization will approach AI. Pretty much akin to what is already being done with sustainability and ESG, make it more official, more formalized. It will be necessary because you’re already seeing For example, the EU, coming up with regulations that will pretty much encourage that if not enforce it. And yeah, I’d say your point about NVIDIA CEO, Jensen Wang’s, notion that could be characterized as, “Mamas don’t let your kids grow up to be coders” is maybe, maybe a bit of an exaggeration in the near term. However, it does point to a key megatrend here, and that is natural language skills that can, assist the prompting can actually be if not more valuable, at least just as valuable as, coding skills as we know them today. That means study up on your English lessons.
Olivier Blanchard: Hey, listen, I’m glad that liberal arts majors will finally get their their moment in the sun. Thanks to thanks to AI prompting. I love this from the side of it and also for the the language and reference side of it. So I’m thrilled.
Ron Westfall: Right on. Right on. Yeah. In the tech world, it gives a little more level playing field for just that. And so that I think is a great point. And with that, I think that does segue into the next area we’re looking at, which is another important alliance when it comes to GenAI and its impact on the telcos. And, the mobile ecosystem that is Amdocs and Microsoft and what they did at the show was that they expanded their partnership to quite simply bring verticalized GenAI capabilities to the telecommunications industry. By combining Amdocs AMAZE platform and Microsoft Azure OpenAI service, and I think we’re going to see more of that. We’re already seeing more of that, but what I think is important is that it’s being targeted specifically at the telecom sector.
Yes, general LLMs, large language models, are important. In terms of, kickstarting, say learning how to use a co-pilot for, many general interest areas, but as we know, at the end of the day, it has to align with the assets, the data assets specifically of the telcos for it to be meaningful in terms of improving business outcomes. And so as such, the duo has established a dedicated lab and a center of excellence to further productize GenAI capabilities with the goal of quite simply accelerating GenAI adoption for the communication service providers. Now, the collaboration is designed to drive tangible business results, as I touched on, by combining deep expertise in the communications industry with GenAI technology, and as such, the industry verticalization can produce carrier grade business outcomes, not just general dialogue that can help say a customer service rep.
But it’s again, organization wide impact, and we’re already seeing that with the BT Group example, that’s not just about low hanging fruit, such as customer support, but many different areas, and so to that end the AMAZE platform is accelerating data ingestion, and enlists, also, I think this is important. And doc’s patent pending. large language model routing capabilities, and that will enable, for example, enhanced model precision through retrieval augmented generation or RAG quite simply and optimize token consumption. And that is important for privacy safeguards, for example.
Now, in addition, Microsoft and Amdocs have constructed the partnership to really boost Amdocs’s customer engagement platform, which includes multiple co-pilot products, including sales, order enrichment, care, proposals and more. And again, furthering that interconnected experience across the entire telco organization. Also, Amdocs is working with, at NVIDIA on optimizing the LLMs to improve the latency, accuracy, and cost effectiveness of the Gen A applications and services. Now, in conversations with CSPs at the show and around the show, I’m seeing that the CSP decision makers are definitely prioritizing data privacy and data security. And so we touched on that already. But these are really. At the top of the list when you’re talking to the telco specifically, and I think that’s gonna be true for many organizations across the board.
But this is how I think the vendors out there need to proceed in terms of, why it’s to help telco GenAI. And this is exactly what the alliance with Microsoft and Nvidia is already addressing. That is, they’re already driving again more effective customer care, but also that process efficiency with the built in safeguards. And so we’re seeing again improvement in areas such as bill experience, improved cell service and also improving internal processes such as configure price quote or CPQ. As we’ve seen, the CPQ process can be complex and also time consuming. And so, through the GenAI CPQ pilot, we see Amdocs customers automating the proposal process.
Olivier Blanchard: Yeah, I think, no, I think what you’re pointing out here is I don’t have a lot of commentary on that because it’s a solid approach. I would just add a little bit of like surrounding color. So one of the things that I’m looking at right now is the expansion of the GenAI space from LLM, so large language models to LMM, which is large mixed models. And so if you understand what large language models. Um, it’s really based on, natural language, right? A mixed model looks at other inputs like images, sound, so music, voice, whatever. And so we’re seeing like an expansion of, GenAI capability space to LMMs. However, in the case of telcos, where you’re looking at internal systems and billings and analysis and different types of processes, most of that for their uses is going to be still based on natural language, whether it’s written or any other form.
And so it makes sense. But then when you’re expanding to other areas of a business, for instance maintenance and repairs and you’re looking at a GenAI to assist maintenance crews and understanding the problem, you may want to have this mixed opportunity or capability where essentially you can take a picture of a problem and the GenAI engine will essentially understand what it’s looking at and can reference user guides or any kind of other mechanical or engineering reference and see what the problem is and give you an answer on that. Or it might be just like a voice problem that’s converted from voice to text and then the input is is analyzed and then an answer can be generated.
So it’s, I think, the LLM approach is like the core baseline, like like year one to year two approach to this, but I would look for this LMM expansion to to take on a little bit more speed as businesses realize that there’s, there are use cases for it and and the service providers realize that, there’s definitely a revenue opportunity there. And it’s going to be handled by the same kind of, it’s still going to be NVIDIA. It’s still going to be essentially the same capability. It’s just a little bit of a different software around it.
Ron Westfall: That’s an outstanding point, Olivier, and I think it can also align with other important capabilities that are being used today, for example, digital twins comes to mind,
Olivier Blanchard: yea,
Ron Westfall: LMM, it would be well suited for that.
Olivier Blanchard: Yep.
Ron Westfall: And also, as GenAI becomes more embedded across, the telco org., I think it’s important that the GenAI co-pilots as well as LMM co-pilots as they emerge, basically aligned with the entirety of, the telco data set. And so what I see are important takeaways is that the telco specific LLM needs to be aligned with just that. It’s better designed versus an off the shelf option. And also that’s also important to, minimize drift or hallucinations, which as we all know, it’s and but I think rounding back to your point, I think that’s an outstanding one, Olivier, about, having low code, no code environments, because again, that will, basically, minimize the need for data scientists and developers to be focused on these types of areas, certainly in the near term. And thus, I think that will quite simply, streamline the automation processes throughout the telco organization.
So that’s really the other shoe dropping in addition to what we already touched on. And that would fully align naturally with the LMM advent. And so with that let’s now turn to, again, the theme of vertical specific and telco specific GenAI and an important announcement that came out at the show. And so let’s look at Qualcomm, speaking of the processor guys, and it’s new Snapdragon X80, 5G modem RF system and its impact on what I call the AI mobile ecosystem. Now, this new offering features new AI innovations based on a dedicated tensor accelerator to improve data speeds, latency, quality of service, coverage, location accuracy, spectrum efficiency, and multi antenna management, as well as power efficiency.
Now, taken together, It’s also a way for basically the telcos to really take that part that we touched on at the very beginning, but this is impacting the 5G capabilities in a very specific and on the ground way, for example, it’s integrating a dedicated 5G AI processor and 5G advanced ready architecture, to achieve what is seen as important milestones, for example really the first 5G modem with fully integrated NBNTN satellite communication support, as well as six antenna architecture for smartphones, 6X carrier aggregation, and AI based millimeter wave range and extension for fixed wireless access CPE. Now, from your view, Olivier, is this really the most advanced modem to antenna platform you’ve seen to date yet?
Olivier Blanchard: Yes, it is. Yeah, no, it is. And the thing is like, love it or hate them, Qualcomm is always going to, well, I can’t say they always will, but historically they’ve come out with with the best connectivity solutions. And it’s fun that, watching them expand into automotive and all these different areas, but this is their core business initially, right. They do modems pretty much better than anybody else, especially this high end not even premium, it’s just flagship stuff. They’re always ahead of everybody. And there’s a good reason why even Apple, despite their, you know, historical legal disputes and whatnot, iPhones are still using Qualcomm modems for the most part and well for the foreseeable future. Yeah, so we, we have this amazing modem RF system from Qualcomm again that sort of like answers all of the engineering questions or tries to address all of the engineering problems or challenges that that 5G and all these devices have. So on the one hand, you need huge throughput that’s reliable and fast with low latency. If nothing else for gaming, Qualcomm is really big with its it’s OEM like partners with mobile gaming and what they call, I think, elite gaming.
And so you need a really powerful modem that can handle all this stuff. Six antennas, it’s massive, you need a lot of AI on board also to manage all that, to handle the the power management. But what I think what’s interesting since your theme is AI, not necessarily all these other applications is A. I. is the number one driver of technology innovation today. And I don’t think there’s any system or system of systems, whether it’s silicon or software or anything between that isn’t designed to either enable more AI capabilities or optimize AI through AI and so what, what’s important to remember here with this modem is if you take a step back and look at the AI ecosystem as like a holistic sort of like a 10, 000 ft view,I’m just using cliches right now, but like, just zoom out, double click circle back. The AI ecosystem is essentially AI in the cloud, right? AI on devices now, or at least it’s starting to be that, and then AI on the edge with like, like on prem and servers and AI capabilities that are like on the edge of the network.
So you have this entire ecosystem of device to cloud, with connectivity in between. And so even though we’re gonna see a lot more AI and especially GenAI workloads happening directly on the device, whether it’s a phone or a P. C. or a tablet or some other device, there’s still a lot of applications that go back and forth and there are gonna be handle on the cloud. There’s a lot of connectivity going on, even if the workload is mostly on the device, there’s still a lot of stuff going on in the background. So you need modem and modem RF systems and SOCs that are designed for this and that are designed for the next five to eight years of this back and forth connectivity of AI in the cloud, AI on device, something in between and all the other processes happening. And in such a way that they’re power efficient, fast, near zero latency reliable, especially with vehicles.
Now that we’re seeing, the automotive space become more connected, have a lot more features inside the vehicles, a lot of GenAI moving into the vehicles as well. You need a modem that can handle all this. And that’s what they have here. And it, I think some of the things that I look at also is Qualcomm’s push of WiFi 7 putting them on cars. And it’s like, you need the modem solution first to be able to have an effective WiFi 7, ecosystem on, on, on the user side of it. There’s also the fixed wireless access piece, which is huge. It’s basically, it’s so much cheaper than laying fiber that essentially wireless fiber is still one of the growth areas, I think, for the industry and to be frank, fixed wireless access is has some engineering challenges and some marketing challenges to overcome.And I think that coming out with a modem that addresses that specifically and helps smooth out the runway for for accelerated fixed wireless access adoption is really smart. So it just it checks all the boxes that a year ago we were talking about it needing to check and it does.
Ron Westfall: Yeah, and I think this announcement really helps tie together. Yes, you definitely need, this particular offering the Snapdragon X80 to address not only emerging new gen AI and AI capabilities, but also really ensure that the ecosystem stays on target, with the progress, it’s already built into the three GPP release 16 through 18 specifications, and I think it’s really a feel good story that, didn’t get a lot of spotlight at the show because, AI was so dominant, is the fixed wireless access aspect. And yes, there is on the horizon, some capacity challenges in terms of the markets that are a little more that are quite simply further along, such as in the U. S. But globally, it’s really a good news story. In fact, it’s one where we can anticipate there will be a tripling of FWA deployments because that’s already proven itself. It’s something that the operators are monetizing successfully. Already there’s proving itself not in terms of just okay, is it a a new way to address digital divide, for example, they’re reaching real customers that have not had broadband yet.
But it’s also proving a compelling alternative to existing say DSL connections, some cable connections. And so there we go. And also, I think, reminds us that 5G advanced, let’s be a little patient here is going to be something that will, I think, really raise 5G’s overall profile in terms of business outcomes and so forth as we patiently see, 5G standalone networks increasingly come on board. And this will be like, the long awaited, “What is next?”, that will, I think, change people’s minds about, “Okay, 5G, it’s a flop. It’s a fizzle.” Well, no, it’s just still a pretty much we’re in the midpoint of the journey and as we know, 5G adoption in its beginning stages happened more rapidly than, say, at the beginning stages of LTE. Yes, I think there was a lot of I guess you could say heightened expectations, but I think, what the show is showing that 5G can certainly align with using GenAI to improve the overall profile of the carrier, but also it’s keeping in line with the natural evolutionary steps that 5G has already presented to the ecosystem.
Olivier Blanchard: Yup, absolutely.
Ron Westfall: And on that, I would say a heartwarming note, thank you so much again, Olivier for joining the call today. I think we were able to both overcome at least, I appreciate you helping with my, this ongoing allergy issues, but that too shall pass.
Olivier Blanchard: Yeah. It’s one of those seasonal things that mine hasn’t even hit super hard yet, but like, I can’t wait.
Ron Westfall: Well, hey I can give you a preview, but with that, thank you everybody for joining and listening in on our podcast. We always appreciate it, and actually don’t forget to bookmark us, reserve us for our next upcoming 5G Factor. And with that, again, everyone have a good and healthy 5G day.
Other insights from The Futurum Group:
5G Factor: Key MWC24 Takeaways – Semis and Devices
5G Factor: Key MWC24 Takeaways – The Cloud and Telcos
5G Factor: Key MWC24 Takeaways – Open RAN
Author Information
Ron is an experienced, customer-focused research expert and analyst, with over 20 years of experience in the digital and IT transformation markets, working with businesses to drive consistent revenue and sales growth.
He is a recognized authority at tracking the evolution of and identifying the key disruptive trends within the service enablement ecosystem, including a wide range of topics across software and services, infrastructure, 5G communications, Internet of Things (IoT), Artificial Intelligence (AI), analytics, security, cloud computing, revenue management, and regulatory issues.
Prior to his work with The Futurum Group, Ron worked with GlobalData Technology creating syndicated and custom research across a wide variety of technical fields. His work with Current Analysis focused on the broadband and service provider infrastructure markets.
Ron holds a Master of Arts in Public Policy from University of Nevada — Las Vegas and a Bachelor of Arts in political science/government from William and Mary.
Research Director Olivier Blanchard covers edge semiconductors and intelligent AI-capable devices for Futurum. In addition to having co-authored several books about digital transformation and AI with Futurum Group CEO Daniel Newman, Blanchard brings considerable experience demystifying new and emerging technologies, advising clients on how best to future-proof their organizations, and helping maximize the positive impacts of technology disruption while mitigating their potentially negative effects. Follow his extended analysis on X and LinkedIn.
