Intel and ASML hit high-volume milestone on High NA EUV, and Intel rolls YOLO26 into OpenVINO

Intel said it and ASML have reached a High NA extreme ultraviolet (EUV) lithography milestone in high-volume manufacturing, a key step in scaling the most advanced tools used to etch ever-finer semiconductor patterns. The update ties the effort to Intel’s Ultra 3 and 18A process node, a next-generation manufacturing platform the company is working to ramp for future products.

According to the market report, Intel characterized yields as being in line with existing manufacturing platforms. That matters because High NA EUV is intended to improve resolution and throughput compared with earlier EUV approaches, but it also raises the difficulty of process development and defect management as fabs move from trials into sustained, production-style output.

The announcement also points to the broader strategy Intel has been describing for its “foundry-like” operations, where the company aims to bring leading-edge manufacturing capability to its own chips and to customers. High-volume qualification is generally treated in the industry as a prerequisite for broader ramp plans, since it indicates that process parameters can be held consistently rather than only succeeding in limited pilot runs.

On the software side, Intel said it is bringing YOLO26 to OpenVINO. OpenVINO is Intel’s toolkit for optimizing and deploying computer-vision and other AI workloads on Intel hardware, including CPUs, integrated graphics, and select accelerators. YOLO, short for “You Only Look Once,” is a widely used family of real-time object-detection models, and adding a new version typically indicates support for updated network architectures and performance profiles that developers want to benchmark and deploy.

While the market report links the hardware milestone and the OpenVINO update in the same news cycle, the two efforts point to different parts of Intel’s roadmap: the manufacturing work focuses on the precision and manufacturability of advanced silicon, while the OpenVINO update targets developer and enterprise demand for AI inference, where tooling compatibility and model support can affect time-to-deployment.

Still, the disclosure in the report appears light on specifics. It does not provide detailed yield figures, ramp timelines, or the number of wafers tied to the High NA milestone, nor does it describe which OpenVINO components will be updated, what optimization paths will be added, or how YOLO26 performance will compare across Intel hardware targets. Readers should treat the news as an operational milestone and a software availability note rather than a quantitative sales or margin announcement.

Going forward, investors and customers will likely focus on whether Intel can translate High NA EUV qualification into broader process ramp execution for Ultra 3/18A, and whether Intel’s developer ecosystem maintains momentum by supporting current AI model versions in OpenVINO. Any follow-on updates that include measurable production metrics, schedule detail, or wider platform support would provide clearer evidence of how quickly these improvements are becoming productized.