NAB Show 2026 has wrapped up at the Las Vegas Convention Center, and for the Scalstrm team it was four days that confirmed what we have believed for some time: the industry is not just asking better questions about video infrastructure, it is actively looking for answers it can implement now.
We were on the floor at the NETINT Technologies VPU Ecosystem Pavilion, from 19 to 22 April. What we brought was not a concept or a roadmap. It was a live production environment, running in real time, demonstrating a different way of thinking about how streaming infrastructure should work.
A Week of Direct and Purposeful Conversations
This year’s NAB Show welcomed more than 58,000 registered attendees from 146 countries, with over 1,100 exhibitors on the floor. The energy reflected an industry navigating significant change, with AI, cloud infrastructure, and the creator economy dominating much of the conversation. But underneath those broader themes, a more specific pressure was present in nearly every exchange we had: broadcasters and streaming platforms need infrastructure that can do more with less. Transcoding, packaging, delivery, without the operational complexity and cost overhead that has become the default.
What Patrik Alm, Ola Bengtsson, Dominique Vosters, and Pontus Eklöf heard consistently throughout the week was a readiness to move. The organisations in those conversations were not evaluating abstract possibilities. They were looking at specific infrastructure decisions, with real timelines, and they wanted to know whether a VPU-powered, software-defined approach could deliver what their existing setups cannot.
What We Demonstrated on the Floor
The live demonstration at Booth W2713 was central to everything we did at the show. Scalstrm’s platform was running as it does in production: every stream generated on request, every frame scheduled and processed through a single coordinated control plane, with no pre-encoded renditions sitting idle in storage waiting for a viewer who may never arrive.
That last point resonates differently when you see it running. The inefficiency baked into conventional approaches, the storage overhead, the compute cycles spent encoding content no one ultimately watches, is something the industry has normalised over time. Showing it working another way, live, made the contrast concrete.
The combination of Scalstrm’s software-defined architecture with NETINT VPU hardware is where the performance and efficiency gains become most apparent. Just-In-Time Transcoding at VPU throughput changes the economics of streaming infrastructure in ways that are not incremental. Storage requirements fall significantly. Energy consumption follows. And the system scales to demand rather than requiring operators to provision for peak capacity at all times.
The Broader Direction of the Industry
What NAB 2026 reinforced is that VPUs are no longer a niche topic. The questions we were fielding on Day 1 were the same ones we were fielding on Day 4. The conversation has matured. Operators are past asking whether VPU-based infrastructure is viable. The question now is how to transition, and what implementation looks like at their scale.
The gains available from VPUs are not automatic. They depend on a platform designed to coordinate across the full pipeline, from ingest through transcoding, packaging, and delivery, without the inefficiencies that accumulate when components are loosely connected rather than purpose-built to work together. That is what Scalstrm is built to do.
Looking Ahead
Thank you to everyone who visited the booth or took the time to meet with us during the show. For those who did not have the opportunity to connect in Las Vegas, we would welcome the chance to continue the conversation. Reach out to the team directly, or visit scalstrm.com to learn more about how Scalstrm and NETINT VPU-powered infrastructure can change what your streaming operations are capable of.
