Summary
NVIDIA and the European Broadcasting Union are expanding a framework for public-service broadcasters to deploy AI while retaining control over data, infrastructure and models. NVIDIA also reports faster video search in Adobe Frame.io workflows used by German producer UFA.
NVIDIA and the European Broadcasting Union (EBU) are expanding a framework to help European public-service broadcasters deploy AI while retaining control over their data, infrastructure, models and security. Separately, NVIDIA says its accelerated-computing technology is helping Adobe Frame.io search video libraries faster, including in workflows used by German television and film producer UFA.
The announcements were presented at a Berlin creative-industry event on October 19, 2026. The EBU initiative is a standards and architecture effort; the Frame.io work applies GPU computing to media processing and search.
A framework for broadcaster control
The EBU and NVIDIA are bringing sovereign cloud providers, open-model developers and systems integrators into standards work that began the previous year. The initiative is intended to define requirements and reference architectures for where broadcaster workloads run, how data is governed, which models are used and how systems are secured through their lifecycle.
The participating organisations describe this as a way to make AI deployment more practical for broadcasters while protecting editorial independence, data sovereignty and national resilience. The EBU represents more than 115 member organisations in over 55 countries, according to NVIDIA.
Microsoft is contributing expertise from Microsoft Sovereign Cloud, including Azure Local and Foundry Local, for local processing of sensitive editorial material and continued operation when external connectivity is unavailable. Nebius offers a platform for training, fine-tuning and running models on European infrastructure. NVIDIA says its Nemotron open models can be adapted for broadcaster-specific applications and public-service requirements. Qvest and other systems integrators are contributing media-sector consulting and implementation expertise.
The work focuses on requirements and reference architectures rather than announcing a single finished system for broadcasters to install. Its aim is to give organisations a shared basis for evaluating how AI services are deployed and governed.
Frame.io search finds moments inside video
Adobe is using NVIDIA accelerated computing in a Frame.io video-search pipeline. NVIDIA says it uses NVDEC for video decoding, TensorRT for optimising AI inference, and CUDA for preprocessing and generating embeddings. Embeddings are numerical representations that let software compare video content with a search query by meaning, rather than relying only on filenames or folder structures.
NVIDIA reports about five times faster embedding generation, up to four times higher throughput and up to five times lower latency for search processing. Adobe is also using NVIDIA NVENC and NVDEC to accelerate video encoding and decoding during media ingest. The figures are presented by NVIDIA as results of the accelerated workflow.
A new proxy format samples video at two frames per second and is about 80% smaller than the 360p version it replaces, according to NVIDIA. Smaller proxies can reduce the amount of data stored or transferred while preserving visual detail needed for search.
The workflow is relevant to UFA, which produces more than 3,500 hours of programming each year, including the daily drama “Gute Zeiten, schlechte Zeiten.” Editors assembling recaps or flashbacks may need to locate a specific reaction or exchange in footage recorded many years earlier. Frame.io semantic search can find moments by concept, action or setting; hybrid search combines that with notes, transcripts and approval status. An editor could search for a character looking concerned, then narrow results by character and approved footage.