Dell Technologies World 2025: Day Two (even more AI, sustainability, dogs, data, …)

Author: Peter Jarich, Head of GSMA Intelligence

The weatherman explained this morning that Thursday is likely to be the first day of the year when Las Vegas registers a 100 degree temperature (Fahrenheit, natch). Luckily, I’ll be on a plane home long before that; several years of living in the UK has made me allergic to balmy weather. 

In the meantime, the 2nd day of Dell Technologies World picked up where the first left off. What does that mean? Think extensions to show themes, things I simply missed yesterday, and a minor degree of randomness. 

• AI – It’s the Same, But Different. You’d expect Dell to position AI as just another IT workload. It’s risen to the top of everyone’s agenda over the past few years, but Dell – or any of their competitors, for that matter – would tell you that it’s an application they can support in the same way they’ve always supported an evolving set of diverse applications. But, across sessions and conversations, there was a clear understanding (from Dell and partners) that AI is different. New silicon needs, sure. But new powering, scalability, cloud, and architecture decisions to be made. Even the silicon question isn’t so clear; some use cases may require GPUs, but many will be fine on CPUs…laptop CPUs, even. In many ways, this mirrors the AI-driven data traffic considerations operators are looking at. Whether or not AI drives a massive amount of new traffic, operators are not sure. But they are much more certain that it will drive traffic patterns.
• Is Data an AI Afterthought? I mentioned yesterday that Day One Dell Technologies World announcements included enhancements to its AI Data Platform. The Dell SME on the show floor was eager to talk about it. Almost too eager. Unlike servers brimming with GPUs or the flashy use case demos, the crowd around the Data Platform pod was…well, it wasn’t quite a crowd. You could, partly, blame Dell for this; the pod was somewhat awkwardly positioned between AI Factory demos vs. integrated into the middle of them. That said, we know that the importance of clean, well-managed data is often not top of mind. For example, in our work with telcos on AI maturity, only 30% claimed Data Quality and Complexity was a top three barrier to AI adoption. Amidst all of the other potential deployment challenges, this might not be surprising. Nonetheless, it’s a clear risk – for telcos and the broader market.
• Sustainability > Efficiency. The default sustainability message for many IT companies is straightforward: our servers are energy efficient. Energy efficiency is clearly part of being sustainable, but only one part. Packaging practices. Use of recycled materials. Product buy-back and recycling. Sustainability services. To its credit Dell’s dedicated sustainability demos had it all. Where they fell somewhat short was a limited focus on the sustainability implications of different network architectures or deployment models. Depending on the use case, some network architectures will be more energy efficient than others. Dell had plenty of stories attesting to this, stretching all the way to running AI models on powerful PCs vs. cloud resources. Quantifying or showcasing those decisions in a simple way – even as a conversation starter for their Services team – would be more powerful.
• Everything’s Better with Dogs. Just before you get to entrance of the Dell Technologies World Expo there’s a fenced in area. At CES, the fences are meant to contain drones. Here, they’re holding back service dogs…and lots of people coming to greet them with ginormous grins. Airports and hospitals long ago recognized that a friendly pup can help to defuse a stressful situation. Tradeshows aren’t quite as stressful, but benefit nonetheless.  
• Smart Cities at the Edge. Another day. Another demonstration of AI at the edge. Unlike the compelling inventory management story from Chooch on Day One, the standout on Day Two – use of AI in traffic management by the City of Bellevue, WA – involved an operator. In this case, AT&T was running AI workloads on Dell servers in its network to process video from traffic intersections in order to improve pedestrian safety. Much like the Chooch example, though, costs were a primary driver of the decision to leverage edge resources; transport costs for massive amounts of video traffic would simply make the use case untenable.