Dust settles after a buzzling telecom week in Barcelona 

The question that arises is "If the surface display of the event matches the business that goes on?" The answer is No. It used to be a show where big contracts were announced, now it is more of an event for networking and showing directions into the future. As for the networking it is a very efficient meeting place as “everybody” goes there. As business is very much about personal relations, MWC keeps its important role, but we tend to take concrete steps after the event. It is also a good place for being creative, mixing ideas and creating new ones. 

So here are some observations (credit to my colleague Johan Forsman for contributions): 

Horizontal telco cloud gains real traction 

Since the inception of NFV more than a decade ago, the vision has been to run telco workloads on a horizontal cloud. A cloud that is more like “a standard” cloud. In the early days that did not fly because of performance reasons. Those issues (and the desire from vendors to sell appliances) led to the vertical cloud approach. I.e. a specialized cloud for each vendor solution, as the only solution possible and one that gives telcos the guarantees they need. Reasons for the initial problems included the lack of cloud nativeness of the workloads. 

This year's MWC saw displays and presentations of some interesting horizontal cloud projects. Firstly, the Linux foundation project Sylva where a multitude of telcos and vendors unify the definition of such a telco cloud that meets needed performance, and scalability. The two most vocal supporters here are Orange and SUSE, but others have followed suit. Tieto is engaged in working with validation concepts stemming from CNTi. 

Another prominent horizontal telco cloud is that from Deutsche Telekom. It is similar to the concept from Sylva and contributions go back and forth. 

Perhaps this trend will lead to the promises of NFV being finally achieved. 

5G is almost here 

Most people think they have been on 5G for a long, long time, and by some definition it is true. The point with this topic is that what is referred to as 5G SA (stand-alone) is now finally starting to be deployed. We, normal consumers, won’t notice the difference in most cases, but for B2B customers this could be a huge difference as 5G SA delivers on the low latency promise through local breakouts, and some of the user privacy claims as the data traffic can be kept local, thereby actually deliver on the original 5G promise. 

AI is everywhere... and nowhere 

The overarching message across this year's MWC is that of AI. It is around AI for telecom operations, it is AI for algorithms in products, it is AI hosted in edge networks (RAN or Metro), it is telecom interconnecting AI workloads to achieve AI performance. 

Everywhere you look, you see an AI message – gets a bit boring if you ask me, as there are fewer real deployments compared to all the buzz. When engaging into discussions, all the expected issues emerge, like how to label all the data and how the model management could look like etc. 

We are still in early days, and the technology is promising, but remember Amara’s law – we tend to overestimate technology impact in the short term but underestimate it long term. 

Some people have proposed the show should change its name from MWC to NWC (Nvidia World Congress). Of course, a joke, but with a fling of reflection as Nvidia is involved in most of what is going on (being in all honey pots). The AI-RAN Alliance continues to push its comprehensive concept of using and supporting AI workloads. We can observe that the concept gets increasingly adopted by major vendors. Tieto is a member of the AI-RAN Alliance alongside the industry’s biggest names.  

There are many proponents of that an AI-native society will require AI inference to be highly distributed. The only actors that can build on this are the telcos. Whether it would require hosting at cell towers remains to be seen as it would require quite massive investments, but it makes sense for telcos to explore their metro locations as potential places to host such workloads (i.e. the network edge). Many speakers said during the show: “it is a good time to be in telecom” 

Digital sovereignty and Cloud 

The political situation currently at hand has brought an increased awareness of the need for sovereignty to Europe. Sovereignty is important for telco operations, of course, but also for Cloud in general. General cloud becomes more prominent at the MWC show, as hyperscalers are there to position their offerings for enterprises in general and telcos in particular. 

Some of the hyperscaler challengers are telcos. Most important is Deutsche Telekom, which of course makes a clear positioning on sovereignty. Whether we will see a shift in market shares for cloud services remains to be seen. But over the last few years the European alternatives have been growing on par with the hyperscalers on the European market. We also have the strategic direction to build a European federated edge cloud continuum. This is something the European Commission has as a strategic priority being developed by the IPCEI-CIS/8ra initiative that led to a launch of a federated edge cloud concept with today five of the most prominent telcos in Europe. Its justification builds on partly the same reasoning as the AI-RAN Alliance, namely that AI inference must be distributed. Whether that holds true or not, we will see. 

Is 6G just around the corner – nope 

Most telco’s have just gone through a major and CAPEX intense 5G rollout (RAN). They want of course to monetize on that, rather than jump on a new bandwagon. The industry now needs to focus on harvesting value from 5G investments, but it does not stop visionary people from pointing out new directions. 

Most telcos expect 6G to be more of a software upgrade. I think it is safe to say that apart from 3GPP specification releases, we will probably see the “G-race” becoming blurrier with a more continuous software upgrade bringing additional capabilities. This is nowadays referred to as the telecom platform. 

The trend of telecom shifting from hardware centric to software centric has been there for the last 10 years or so, and it is accelerating. Here Tieto is well positioned. 

Autonomous networks 

As cost focus increases in the telecom industry, the idea of autonomous networks gains increasing interest. This is normally fuelled by advances in AI technology. It is however a field in where the gap between CTO visionaries and real operations is quite huge.  We can see a race in claiming the highest score on the TMF levels. 

The industry tends to look at AI as the technology to solve all problems, when in fact we should be searching for the most appropriate solution for each problem. Quite often it could be AI, but not ML (and sometimes ML). In other cases, it could be algorithms or “normal” intent-based automation. This requires an overarching architecture in where to evolve these technologies.  

Tieto wrote a blog post on the topic a year ago. It is still valid, I think: The role of AI in creating autonomous networks. 

We have some interesting offerings together with our partner Inmanta in this area, which I recommend to check out. 

With these thoughts, I am wrapping my observations on this year’s MWC – the time would surely fly until we meet again in Barcelona next year!