Enterprise networking has hit a turning point — it’s no longer judged by port speeds but by how well it keeps artificial intelligence running across clouds, data centers and the edge.

As MWC Barcelona 2026 approaches under the banner of “The IQ Era,” the transition is unmistakable: Intelligence is no longer layered on top of connectivity — it is redefining the purpose of infrastructure itself. This evolution is changing how infrastructure is bought, built and judged. Operators are moving from throughput upgrades to outcome-driven design, and it’s why the networking story at MWC will feel more concrete than it did a year ago, according to Dave Vellante, chief analyst at theCUBE Research, in a Breaking Analysis episode.

What’s different this year is the expectation of measurable AI outcomes. Enterprises are no longer evaluating infrastructure on speeds and feeds alone; they are assessing how effectively networks enable GPU utilization, reduce inference latency and maintain governance across distributed environments.

Join theCUBE, SiliconANGLE Media’s livestreaming studio March 2-5 for exclusive coverage of MWC Barcelona. Interviews will explore how enterprise networking is being redesigned to support AI-driven workloads, distributed infrastructure and measurable performance outcomes. (* Disclosure below.)

Enterprise networking and the AI factory era

AI infrastructure is pushing enterprises to treat the network as part of the compute stack, not an afterthought. At MWC, networking decisions will be tied directly to GPU efficiency and distributed inference performance. Companies such as IBM, Intel, Cisco, Wind River, the Netherlands’ Ecosystem Services and Arm are part of that shift, where fabric design and automation now matter as much as bandwidth, according to Bob Laliberte, principal analyst at theCUBE Research.

“When we look back on 2026, I really believe we’re going to remember this as the year that network stopped being AI adjacent and became structurally essential to AI success,” he said. “I think this is the year that the network really has more of a determining factor for AI outcomes. Vendors need to translate a lot of that AI complexity into control, into trust and measurable business results that will really help define that next era of enterprise networking.”

Enterprise leaders should look beyond raw bandwidth, he added: “Connectivity is no longer just about bandwidth. It’s about deterministic performance, isolation, sovereignty and predictable outcomes tied to business KPIs.”

The practical takeaway is that enterprises are increasingly looking for proof that networks can deliver consistent results across shared environments. That becomes more important as AI workloads stretch across private data centers, colocation facilities, edge locations and multiple clouds, where consistency and isolation are harder to maintain. In that landscape, performance can’t simply be assumed — it has to be engineered, monitored and continually validated, according to Scott Hebner, principal analyst at theCUBE Research.

“Once they did that, what happened was there was latency and performance issues, and it was those businesses that really doubled down on networking to create that great experience that ended up winning,” he said. “That’s what’s going to happen with AI. It needs that networking power to operate as time goes by.”

What networking teams will need to prove

The networking conversation is moving toward operational accountability: how quickly training completes, how efficiently GPUs are fed, what inference latency looks like at the edge and whether segmentation and governance can hold up when tenants, models and data pipelines collide. That’s a different buyer mindset than “refresh the switches,” and it changes what vendors have to show in Barcelona, Laliberte explained.

“It won’t be measured from a who’s the first person to get to 1.6 or this that other terabit throughput,” he said. “It’s going to be have the combined solution from an end-to-end that will really drive these outcomes for AI.”

Several themes should dominate the conversation in Barcelona, according to Laliberte: deterministic performance for AI workloads, AI-driven network operations and agentic automation, WAN architectures as part of the AI supply chain, sovereignty and governance embedded into infrastructure design, and outcome-based metrics replacing raw throughput as buying criteria.

For MWC audiences, the useful lens is to watch for where networking is becoming the enforcement point for policy and trust, not just a carrier of packets. When AI workflows become more autonomous, the network has to support that autonomy without letting control disappear. That’s where resilience, segmentation and governance stop being overlays and start becoming intrinsic, according to Laliberte.

“I think next the network’s role is going to be expanding the data center,” he said. “AI workloads are just inherently more distributed, making that latency jurisdiction and bandwidth elasticity strategic variables rather than just operational details. Think about retail, healthcare, industrial environments, all the autonomous robotics, and things like that that are going on all requiring some second decision-making. All those become primary metrics of success. Across all the domains, security, governance and enforcement move from overlays to intrinsic capabilities and the network becomes the control plane for trusted AI execution.”

The Wide-Area Network is taking on a more strategic function, according to Laliberte, who noted that as AI workloads spread across environments, the WAN effectively becomes embedded in the AI supply chain, where latency, regulatory boundaries and bandwidth flexibility carry real business consequences.

The overarching shift is clear. Infrastructure conversations are moving from capacity to capability and from technical specifications to measurable business impact.

“The vendors that translate AI complexity into control, trust and business impact will define the next era of enterprise networking,” Laliberte said.

TheCUBE event livestream

Don’t miss theCUBE’s coverage of MWC Barcelona March 2-5. Plus, you can watch theCUBE’s event coverage on demand after the event.

How to watch theCUBE interviews

We offer you various ways to watch theCUBE’s coverage of MWC Barcelona, including theCUBE’s dedicated website and YouTube channel. You can also get all the coverage from this year’s events on SiliconANGLE.

TheCUBE podcasts

SiliconANGLE’s “theCUBE Pod” is available on Apple Podcasts, Spotify and YouTube, which you can enjoy while on the go. During each podcast, SiliconANGLE’s John Furrier and Dave Vellante unpack the biggest trends in enterprise tech — from AI and cloud to regulation and workplace culture — with exclusive context and analysis.

SiliconANGLE also produces our weekly “Breaking Analysis” program, where Dave Vellante examines the top stories in enterprise tech, combining insights from theCUBE with spending data from Enterprise Technology Research, available on Apple Podcasts, Spotify and YouTube.

Guests

During MWC Barcelona, theCUBE will feature interviews with leaders from across the enterprise networking ecosystem, including Wind River, Netherlands (Ecosystem Services B.V.), IBM and others. Conversations will explore how organizations are redesigning networks to support AI factories, distributed workloads and real-time inference — and how connectivity is becoming central to performance, resilience and measurable business outcomes at scale.

(* Disclosure: TheCUBE is a paid media partner for MWC Barcelona. Sponsors of theCUBE’s event coverage do not have editorial control over content on theCUBE or SiliconANGLE.)

Image: SiliconANGLE