Hybrid Cloud, AI, and Quantum Threats: How Telcos are Rebuilding the Network Edge for Tomorrow

Your Network’s Edge: The RAD Way Episode 4

Overview

RAD Podcast

In this episode of Your Network’s Edge: The RAD Way, we explore the evolving landscape of business cloud access and telco cloud services. We’ll look at the shift towards hybrid cloud architectures and the increasing demand for direct, managed access to cloud resources, driven by digital transformation and the burgeoning impact of AI.

The podcast discusses critical considerations for businesses accessing cloud services, including performance, security, data sovereignty, and cost control, while also examining how telcos are adapting their networks to offer enhanced cloud connectivity. Furthermore, we share insights from a survey of telecom professionals, revealing their motivations for adopting telco cloud services and the growing importance of network edge clouds to meet the specific requirements of modern business applications.

 

Follow the Your Network’s Edge: The RAD Way podcast on Spotify,  Apple Music, or YouTube. Or click the play button in the embedded player below:

Transcript

Introduction – Hybrid Cloud & AI in Enterprise Connectivity

0:00 Welcome to the deep dive. 0:00 Today we’re uh diving head first into how businesses are connecting to the cloud and maybe more importantly why those you know invisible threads of telecommunication networks are now more critical than ever before. 0:12 We’re seeing this fundamental reshaping really of enterprise connectivity. 0:16 It’s driven by relentless digital transformation and well let’s face it the huge rise of AI. 0:23 Exactly. 0:23 And this deep dive is really all about understanding the powerful um managed carrier networks underneath it all, the ones providing that essential access to these crucial cloud services. 0:35 We’ll explore the new challenges, sure, but also the incredible opportunities this brings for businesses and for the Telos connecting them. 0:40 And to help us navigate this uh rapidly changing landscape, we’ll be drawing on some unique insights from a recent independent survey. 0:46 It focuses specifically on telco cloud access. 0:51 Right, our mission today is simple. 0:51 cut through the jargon, reveal the key trends, the challenges, some are pretty complex, and also some genuinely innovative solutions emerging in this vital area. 1:01 We want to help you get well informed quickly. 1:03 You know, those aha moments that just make sense of it all.

Hybrid Cloud Basics & East–West Traffic Explained

1:09 Okay. 1:09 So, to really grasp this evolving landscape, we have to start with well, the cornerstone of modern business connectivity, the hybrid cloud. 1:20 It’s become the deacto standard, hasn’t it? 1:21 It’s no longer just about using one public cloud or, you know, sticking only to your own private data center. 1:27 Instead, you’re seeing businesses strategically spreading their applications, their data, which is precious, and their compute resources across multiple public clouds, private clouds, and yes, still their own private data centers. 1:37 It’s a truly distributed digital footprint. 1:40 And this while this has profound implications for the telco network, essentially the telco network is now inheriting what used to be kind of confined within a data center network. 1:48 This leads to entirely new traffic paradigms. 1:49 like we’re seeing a significant jump in what’s called east west traffic. 1:52 East west traffic. 1:52 Okay, explain that a bit. 1:56 Think of it like this. 1:56 Traditional north south traffic is data moving in and out of your main data center or maybe a cloud, right? 2:04 East west is data moving sideways. 2:04 So from one application to another within the cloud or maybe even between different clouds or regions. 2:10 It’s like the internal chatter between systems and it’s exploding as apps become more distributed. 2:17 Got it. 2:17 So it’s not just about getting to the cloud anymore. 2:20 Exactly. 2:20 It means telos are connecting not just business sites to the cloud but increasingly data center sites to cloud locations as well to enable that critical intercloud communication.

AI’s Impact on Data Traffic and Physical–Digital Convergence

2:31 Okay, we’ve talked hybrid cloud implications but what happens when you throw AI into this mix? 2:35 I mean how does that specifically amplify these network demands? 2:39 What does it mean for data traffic? 2:41 Oh, it’s a massive multiplier, a huge one. 2:43 The entire AI life cycle, right? 2:45 from collecting raw data to model training and then that continuous inferencing, it introduces even more of this east west traffic. 2:51 And if you connect that to the broader picture of digital transformation, well, it’s really about capturing physical spaces. 2:58 Physical spaces like what? 3:00 Think of manufacturing workshop floor or complex hospital operations and bringing them into the digital realm. 3:07 This happens using you know tons of IoT sensors, cameras constantly streaming video, machines reporting operational data. 3:14 Yeah. 3:14 all that stuff. 3:14 Then AI takes this digital picture and reasons feedback back to the physical world, creating this incredibly powerful closed loop system. 3:26 Wow. 3:27 And this dynamic, as you can probably imagine, is driving just exponential data growth across the network. 3:33 And we’re talking specific use cases here that really push the network, aren’t we? 3:36 There’s a whole life cycle of distributed data, things like uh migration, sync, backup, recovery, all happening constantly. across these different cloud locations. 3:44 Then you’ve got access to elastic compute. 3:46 Imagine a business suddenly needing to burst to cloud compute for peak loads or maybe accessing remote GPUs for special tasks. 3:54 And this is of course profoundly affected by AI which demands just massive ondemand processing power. 4:00 And that brings us to the AI life cycle itself. 4:02 Huge data ingestion from devices, sensors, cameras. 4:04 You’ve got federated training, distributed training, even augmented generation pulling data from remote LLMs. 4:13 Yeah. 4:13 Yeah. 4:13 And what’s fascinating here is how these new use cases dictate new characteristics for the data traffic itself. 4:18 For example, AI training can mean these massive bursts of data running for hours straight followed by periods of well relative calm, very bursty, right? 4:27 Unpredictable. 4:29 Exactly.

Bursty Traffic Patterns & The Need for Deterministic Performance

4:29 And we’re also seeing more upstream traffic. 4:30 Think endpoints sending continuous video streams up for AI analysis. 4:34 All of this demands a critical new requirement. 4:36 Deterministic performance. 4:39 Deterministic performance. 4:39 Okay, that sounds important. 4:41 What does it really mean in practice? 4:44 Imagine, say, a realtime medical procedure being guided remotely or maybe an autonomous vehicle reacting to its environment. 4:50 A tiny unpredictable delay or a dropped packet. 4:53 It can be catastrophic. 4:55 Deterministic performance ensures data arrives not just quickly, but consistently within precise guaranteed time frames. 5:01 It makes the network behave less like a best effort public road and more like a precisely timed dedicated railway line. 5:08 Ah okay that makes sense. 5:11 Predictability is key. 5:13 It’s a level of network predictability that you know used to be reserved only for mission critical industrial systems. 5:18 Yeah. 5:18 But now is becoming expected for everyday business operations. 5:22 So okay we’ve established this critical need for deterministic performance. 5:25 But what other non-negotiables are businesses really pushing for in this new cloud era? 5:31 What else is top of mind?

Security, Data Sovereignty & End‑to‑End Observability

5:33 Well security is always always a primary concern. 5:35 That hasn’t changed. 5:35 But increasingly because of things like data sovereignty and privacy regulations, businesses now deeply desire to control the movement of their data. 5:44 Uh right, GDPR and things like that. 5:46 Exactly. 5:46 This isn’t just about ticking a compliance box. 5:48 It’s a significant new focus on having granular control over where data resides and where it travels. 5:56 They also want endto-end observability. 5:59 They need to see what’s happening across this whole distributed hybrid cloud network so they can quickly troubleshoot problems. 6:04 It’s about gaining clear visibility into what’s become a very complex environment. 6:10 Makes sense. 6:10 And what we’re also hearing, I think, is that businesses want these services, networking included, delivered as a service available on demand. 6:16 You know, much like how they consume cloud compute today. 6:20 Yes, that flexibility is key. 6:22 And crucially, they absolutely want the ability to control and hopefully optimize what can become really complex network and cloud service costs. 6:28 Because let’s be honest, cloud network services, they are not cheap. 6:33 Those costs can quickly spiral if you’re not managing them properly. 6:37 Definitely cost control is a huge factor.

Evolution of Connectivity – Internet, SD‑WAN & Secure Service Edge (SSE)

6:39 Okay, so maybe let’s trace the journey of cloud connectivity for a moment. 6:44 Where did this all start? 6:46 It really began with the internet providing just you know universal access and for a long time that was kind of enough. 6:53 Then as businesses adopted more cloud services solutions emerged right 6:58 SD1 came along. 6:58 These were designed to enhance performance and reliability over regular broadband in the internet. 7:05 They brought things like automation, better observability, service awareness. 7:08 It was like upgrading from, I don’t know, a local road to a slightly more optimized highway, maybe. 7:15 Yeah, that’s a good analogy. 7:16 And that evolution continued, didn’t it? 7:16 SD1 was joined by secure service edge or SSE. 7:22 This was a major shift, moving the security perimeter itself from individual business sites to a cloud hosted location. 7:27 Why was that so important? 7:30 Well, it was essential for adhering to that multihyper cloud architecture we talked about and also supporting the growing number of remote workers needing secure access from anywhere. 7:39 It basically acknowledged that the old castle remote security model just didn’t work anymore when your castle was effectively everywhere, right? 7:44 The perimeter dissolved. 7:46 Okay.

Direct Cloud Access – Private, High‑Performance Connections

7:46 And this brings us then to the next major step, direct cloud access. 7:50 We’re now seeing businesses actively looking for direct cloud access with you know the utmost performance and crucially that deterministic performance we discussed earlier. 7:59 So what exactly is direct cloud access? 8:02 How does it differ fundamentally from say just an enhanced SD1 connection over the public internet? 8:09 Okay, so direct cloud access offers truly private dedicated access. 8:14 It’s established directly between your network and the cloud provers network. 8:18 Critically, it bypasses the public internet altogether. 8:19 bypasses the internet. 8:21 Okay, think of it like having your own dedicated secure express lane directly into the cloud provider’s data center instead of using that shared public highway with everyone else. 8:31 It’s often called things like direct connect or express route by the hyperscalers. 8:36 Ah, I’ve heard those terms, right? 8:37 And this service gives you private access over the telco network, not just to your private business workloads in the cloud, but also to public cloud resources like specialized storage or high performance GPUs. 8:50 It’s implemented in a couple of main ways. 8:52 Either the telco can host its own demarcation device, basically a specialized network box right at the hyperscaler location. 9:00 Okay, that marks the boundary where the telco network connects directly into the hyperscaler’s infrastructure. 9:05 It aggregates traffic from different business sites to offer one unified direct cloud connect service. 9:08 Yeah. 9:10 Or alternatively, the business customer might host their own router at the hyperscaler location and then use a layer 3 or maybe a layer 2VPN service provided by the telco to get that access. 9:21 Got it. 9:21 But the key takeaway here and it seems like a really critical one is that in both scenarios, the access network bypasses the internet entirely. 9:29 So this provides a fully managed private access service. 9:31 It ensures the required performance levels are met and importantly gives businesses that precise control over data movement. 9:39 Unlike traffic just flowing over the public internet, it’s about taking that variability out of the equation.

Managing Performance – End‑to‑End Visibility and Troubleshooting

9:44 Precisely. 9:44 And this managed underlaybased access. 9:46 It’s being further enhanced now for better observability and troubleshooting. 9:51 Solutions are now providing end-to-end serviceaware observability. 9:55 They do this by dividing the connectivity into three distinct domains. 9:58 the local area network, the direct cloud access link itself, and then the cloud domain on the other side. 10:07 How do they do that? 10:08 It’s typically achieved by intelligently collecting and analyzing traffic samples at key points. 10:11 This gives you a clear, granular picture of performance across that entire chain. 10:15 So you can pinpoint exactly where a problem might be rather than just knowing, you know, something is slow somewhere. 10:22 That level of visibility must be invaluable.

Quantum‑Safe Encryption & The “Harvest Now, Decrypt Later” Threat

10:26 Okay, moving to another critical area. 10:28 Let’s revisit SSE that secure service edge as the sort of de facto security architecture. 10:33 We talked about how all business traffic gets on-ramped through this cloud hosted security perimeter applying zero trust principles other cloud access security measures. 10:44 But what else is on the horizon especially when we think about really futureproofing our data security? 10:49 Well, something that’s truly fascinating and frankly quite unsettling is the emerging discussion around quantum safe encryption. 10:55 Quantum safe encryption. 10:55 Okay. 10:55 The risk is very real. 10:57 Quantum computers are developing incredibly rapidly. 11:02 Soon they’ll likely be available as a service from cloud providers. 11:03 And these machines could potentially break common network encryption methods we rely on today. 11:09 Things like RSA based IPSec, MAXSE, MQTT, TLS potentially in a matter of hours. 11:15 Hours. 11:15 Wow. 11:16 Yeah. 11:16 And this isn’t just a problem for, you know, some distant future. 11:18 Hackers are already thought to be employing Harvest now decrypt later attacks. 11:23 Harvest now decrypt later. 11:23 What’s that? 11:25 They’re storing encrypted data today with the full intent to decrypt it in say 3 to 5 years when quantum computers become more common and powerful enough. 11:36 This means your critical business data harvested today could be completely compromised down the line. 11:41 That’s genuinely chilling. 11:41 So what does this mean for you know the average business with sensitive data? 11:45 What can they even do? 11:47 Well, carriers are already looking into upgrading their infrastructure to be quantum safe and they’re starting to offer quantum safe encryption services to businesses often through solutions like postquantums. 12:00 What this signals really is that businesses need to start thinking about crypto agility. 12:05 Crypto agility. 12:06 Yeah. 12:06 Designing their systems so they can easily swap out encryption methods as new quantum safe standards emerge rather than, you know, waiting for some kind of quantum computing Y2K moment to hit 12:17 that harvest. 12:17 now decrypt later concept is really unsettling. 12:19 Paints a stark picture of proactive attackers.

DDoS in the Cloud Era – Layer 2 Connection Security Challenges

12:24 Okay. 12:24 But shifting to another cloud security risk, one that’s maybe more immediate for many. 12:27 DOS attacks, distributed denial of service. 12:30 Cloud connectivity offers huge benefits. 12:32 Sure. 12:35 But it can also be leveraged for attacks targeting business sites, right? 12:39 How are carriers adapting to mitigate that threat? 12:40 Yeah, this presents a significant challenge because those traditional Dto scrubbing centers, they kind of become irrelevant for certain types of cloud access, especially layer 2 direct cloud access services. 12:52 And why irrelevant? 12:53 Well, imagine traditional scrubbing centers like a big detour system. 12:57 They divert all your traffic to a security checkpoint, clean it up, and send it back. 13:00 That works okay for layer 3 internet traffic where data gets routed like mail with an address. 13:06 But with direct cloud access, which often operates at layer 2, it’s more like a private sealed tunnel directly to the cloud. 13:12 There’s no easy detour point built in. 13:15 Oh, okay. 13:15 So, the old security checkpoint can’t easily intercept the traffic in that private tunnel. 13:18 Exactly. 13:21 It means telos need new, smarter ways to spot and stop attacks within that private tunnel itself. 13:26 So the augmentation solution being offered now involves collecting traffic samples directly from the network devices within that path and then applying AI based thread intelligence to detect and mitigate these DOS attacks. 13:40 It’s effectively adapting security to the specific nature of these layer 2 direct cloud connections.

AI‑Powered DDoS Detection and Mitigation

13:45 AI fighting AI in a sense. 13:45 Okay. 13:48 Given everything we’ve discussed, the demands, the challenges, these new solutions, it really begs the question, what are the Telos themselves actually seeing in planning? 13:54 We’ve got some really illuminating insights here from that recent independent survey. 13:57 We mentioned 250 senior telecom pros, tier one telos across North America, Europe, Asia, Pacific. 14:03 It gives us a great pulse on where the industry sees itself heading. 14:07 Indeed. 14:07 And we looked at their motivations for operating a telco cloud in the first place. 14:10 Roughly 24% are primarily focused on strengthening their own infrastructure using virtual networking, hosting virtual network functions, that sort of thing. 14:20 A larger group, 29% plan to offer managed services, especially data services, clearly aiming to move up the value chain beyond just connectivity. 14:28 Makes sense. 14:28 Offering more value. 14:30 Then you have 25% who are evolving to become network edge cloud providers for their business customers and another 22% are focused on hosting thirdparty providers on their infrastructure. 14:41 So it shows a diverse set of strategies, doesn’t it? 14:44 It does. 14:44 But the underlying insight I think is that telos are now battling for different pieces of the overall cloud pie. 14:50 Some are doubling down on foundational infrastructure, others aiming for those higher value managed services. 14:55 And a significant portion are pivoting to become the edge itself. 15:00 This isn’t just about offering cloud access. 15:02 It feels like they’re redefining their core business model for the digital age. 15:05 That is a powerful shift in their identity.

Telco Competitive Edge – The Importance of the Network Edge

15:07 So building on that, why do Telos believe businesses will actually choose their telco cloud services over say going straight to a hyperscaler or using other options? 15:16 What’s their perceived competitive edge? 15:19 Well, about 19% cited access to managed services as the key driver. 15:22 Another 26% mentioned attractive pricing bundles, you know, combining network security and cloud. 15:29 But here’s where it gets really interesting. 15:31 A significant 55% over half believe their physical location at the network edge is a key differentiator 15:38 edge location itself. 15:39 Yes, they see it as enabling that managed access with lower latency by bypassing the internet which as we’ve discussed is a huge deal for deterministic performance and crucially 42% believe that rising AI demands will inevitably shift business workloads from on-remise servers to the telco edge cloud. 15:57 Why? 15:57 because businesses will seat scalable resources closer to their actual premises. 16:03 That’s a compelling vision. 16:04 Telos becoming this new hyper local extension of the cloud. 16:07 That 55% figure really underscores the strategic importance of the edge for Telos. 16:14 Okay, so looking at the flip side, what are the current managed access methods businesses are actually using today according to this survey? 16:20 What does the data show? 16:22 Right, looking at the current methods, the survey data shows 62% are using direct cloud access already. 16:26 59% are using thirdparty SD1 services and 34% are opting for their service provers managed SD1 offering. 16:34 Additionally, 40% are still using traditional layer 2 and layer 3VPN. 16:40 So quite a mix still 16:41 of definite mix. 16:41 But what’s truly fascinating here is that among those still using layer 23 VPNs, a notable 21% about one in five are already recognizing the importance of deterministic networking. 16:51 They’re specifically citing the need to bound latency and jitter for their telco cloud access. 16:56 Even with older VPN tech, they see the need. 16:59 Yes, it highlights this growing awareness of these really stringent performance requirements. 17:03 It indicates the market isn’t just asking for better networks anymore. 17:07 It’s demanding a level of precision that again was once only reserved for very specific missionritical systems. 17:15 It’s like a silent signal that the future of enterprise applications is just inherently tied to network predictability.

Closing Summary & Future Outlook on AI‑Driven Connectivity

17:21 And it all loops back, doesn’t it? 17:22 back to what we started with digital transformation, the rise of immersive applications like VR and AR. 17:28 These truly demand those deterministic networks, you know, low loss, bounded latency, bounded jitter, particularly right there at the telco edge to really function effectively to come alive. 17:39 Okay, so to quickly recap our deep dive today, we’ve explored the really significant shift towards hybrid cloud architectures. 17:45 We’ve seen the profound and growing impact of AI on network traffic patterns and the escalating demands from businesses for both robust security and that granular control over data movement. 17:56 We’ve also seen how Telos are actively responding to these challenges with innovative solutions like direct cloud access and by strategically positioning themselves with edge cloud services. 18:06 They’re effectively reinventing their role in this whole digital ecosystem. 18:09 Yeah, and as we always try to emphasize, knowledge is really most valuable when it’s understood and then applied. 18:14 Consider how this evolving network infrastructure isn’t just, you know, a technical upgrade. 18:19 It’s a fundamental enabler for the next wave of digital innovation. 18:23 We’re talking about everything from highly efficient industrial IoT deployments to truly immersive VR and AR experiences that are poised to reshape entire industries and how we all interact with technology. 18:37 Which brings us to a final provocative thought for you, our listeners, to consider. 18:40 With this increasing convergence in the physical and digital worlds and this really urgent looming need for quantum safe encryption, how might businesses need to rethink their entire data strategy? 18:50 Not just their network, but their whole approach to data to truly futureproof their operations in a world that’s going to be increasingly driven by AI and ubiquitous high performance connectivity. 19:00 Something to maul over.