embedUR systems announced a new partnership with Alif Semiconductor® to deliver native support for Alif’s Ensemble Series of microcontrollers within ModelNova Fusion Studio, embedUR’s end-to-end Edge AI development platform. The integration introduces the industry’s first desktop development environment to provide a complete, train-to-deploy workflow specifically targeting microcontrollers with built-in hardware acceleration for transformer neural networks.
The collaboration closes a long-standing gap in Edge AI development. Today, engineering teams often spend months stitching together open-source frameworks, vendor-specific compilers and fragmented deployment utilities. Fusion Studio brings the full Edge AI lifecycle into a single desktop application-from dataset preparation and model training to benchmarking and on-device deployment-helping teams eliminate cloud dependency, reduce operational costs and simplify complex toolchains.
Alif’s Ensemble Series is the first commercially available microcontroller family purpose-built to enable generative AI and transformer-based workloads directly at the endpoint. The second-generation Ensemble E4, E6 and E8 devices deliver more than 450 GOPS of compute performance, supported by 9.75MB of on-chip SRAM and 5.5MB of MRAM. This architecture enables on-device execution of compact language models and transformer-based vision models that were previously impractical for battery-powered and resource-constrained systems. With an integrated image signal processor supporting up to 60 frames per second at 2MP resolution, developers can run computer vision, traditional machine learning and transformer models concurrently on a single device.
Also Read: OMRON and KST Technologies Partners to Develop LoRa Module for Weather Sensor
At launch, Fusion Studio’s Ensemble support will focus on computer vision workloads, including object detection, pose estimation and image segmentation. Additional AI modalities are planned throughout 2026 and into 2027, expanding support across audio processing, sensor-driven applications and emerging generative AI use cases at the edge.
To showcase the new workflow, embedUR and Alif Semiconductor will host a joint technical webinar on March 4, 2026, demonstrating the complete pipeline from dataset preparation through on-device deployment. Registration information is available for participants who wish to attend virtually.
The companies will also present live demonstrations at embedded world 2026, taking place March 10–12 in Nuremberg, Germany. Visitors to Hall 4, Booth 600 will see real-time deployments of Edge AI models created in Fusion Studio and running on Alif Ensemble hardware.
“Alif’s Ensemble Series represents a fundamental shift in what’s possible at the edge, and Fusion Studio is the first development platform built to match that capability,” said Rajesh Subramaniam, CEO & Founder of embedUR systems. “With Fusion Studio, developers go from unlabeled datasets to running inference on production-class hardware in a single workflow — no cloud costs, no toolchain gaps, and no guesswork.”
“When Alif introduced Ensemble in 2021 as the first general purpose MCU family with support for hardware accelerated AI-inferencing, we immediately started getting questions from customers about tools and enablement for using these capabilities in a streamlined manner.”, said Reza Kazerounian, President & Co-Founder, Alif Semiconductor. “Support in embedUR’s ModelNova Fusion Studio is an important piece of that puzzle, it sets our users up to be productive immediately, and shorter development cycles means quicker time to market. A win-win for everyone.”
