Insider Brief
- microagi formalized a compute partnership with Google Cloud and Nvidia to support Atlas, its data and deployment platform for industrial robotics.
- The Munich-based company said Atlas now trains and runs models on Nvidia Blackwell infrastructure through Google Cloud, along with Google Cloud’s Gemini Enterprise Agent Platform and AI models for multimodal data.
- microagi said the partnership is intended to lower robotics training and operating costs while helping European industrial customers process physical AI data on European compute capacity.
microagi announced it is formalizing a compute partnership with Google Cloud and Nvidia to support Atlas, its data and deployment platform for industrial robotics.
According to the Munich-based robotics company, Atlas is now training and running models on Nvidia Blackwell infrastructure through Google Cloud. The setup includes GB300 NVL72 rack-scale systems, A4X Max instances, RTX PRO 6000 Blackwell GPUs and G4 virtual machines, along with Google Cloud’s Gemini Enterprise Agent Platform and AI models for processing video and other multimodal data.
Atlas is designed to help industrial companies collect factory data, train robotics models and deploy those models into production environments. microagi said the compute partnership is meant to address one of the harder problems in robotics of turning physical-world data from live production sites into models that can operate robots in specific plants.
“Robots learn from data, and data is useless without the compute to train on it,” microagi founder and CEO Bercan Kilic said. “This partnership gives us scale, and it gives us engineering depth we could not build alone at this stage: Nvidia optimises down to the card, Google Cloud across the whole machine. Our customers see the difference in what a deployment costs.”
The partnership comes one week after microagi announced what it called the largest seed round in German history. The company said engineering teams from microagi, Google Cloud and Nvidia are working directly together on the infrastructure behind Atlas.
The company pointed out that robotics training creates uneven demand for computing power. A deployment may begin with months of data collection at a customer site, followed by a short period when that dataset is ready and a model needs to be trained for that facility. At that point, compute needs can rise sharply.
What the Partnership Means
microagi said Nvidia engineers are working on GPU-level performance, while Google Cloud engineers are supporting the wider cloud stack, including virtual-machine configuration, CPU use and job orchestration. Since the collaboration began, microagi said it has roughly doubled the amount of work it gets from each unit of energy.
The company said those lower training and operating costs could make automation more practical for industrial customers, particularly in Europe, where manufacturers are under pressure to improve productivity.
“Robotics is becoming one of the most demanding frontiers for AI, requiring massive physical-world datasets, accelerated compute and a full-stack platform to turn models into intelligent machines,” added Tobias Halloran, director of EMEAI startups at Nvidia. “By running on Nvidia Blackwell powered instances on Google Cloud, microagi can scale the training and deployment of embodied AI systems for commercial and industrial environments.”
Most of microagi’s training runs on Google Cloud capacity in the Netherlands. The company said it requests European compute capacity for European customers as a matter of policy.
“We support microagi with compute, with the Gemini Enterprise Agent Platform and with leading models for processing multimodal data,” noted Marianne Janik, Ph.D., VP EMEA North, Google Cloud. “Just as important, we deliberately reserve compute for startups in Europe, because access to compute decides whether Europe’s data can be processed in Europe. Physical AI is an area where Europe, and Germany in particular, has developed real ambition.”
Europe in Focus
microagi indicated the partnership is part of a larger European industrial issue. According to the company, Europe has extensive manufacturing knowledge but limited control over the chips and hyperscale cloud capacity used by its own companies.
The company said demand from European robotics and industrial customers could help expand regional compute capacity. It pointed to the gap between China and Europe in factory-robot installations, saying China installed 295,000 factory robots last year while Europe installed 85,000.
“The market is usually fair: where there is demand, there is supply,” Kilic explained. “All we are trying to do is create more demand for European compute, because the industrial base we are trying to rescue is European. If Europe does not start buying and building serious compute in the next eighteen months, the gap stops closing.”
microagi said industrial companies interested in Atlas can contact the company through its Atlas contact form.
