Mythic has acquired Videantis GmbH in a move aimed at accelerating development of energy-efficient hybrid AI computing platforms for data centers, automotive systems, robotics, and defense applications.
The acquisition combines Mythic’s analog compute-in-memory technology with Videantis’ digital processor architecture and software stack. Mythic said the combined platform is designed to deliver up to 100 times greater energy efficiency than conventional GPU-based AI systems.
The announcement follows Mythic’s recently disclosed collaboration with Honda to co-develop next-generation AI chips for future vehicles. The company said the Videantis acquisition expands its ability to scale AI computing platforms across applications ranging from drones and robotics to autonomous systems and data center infrastructure.
The deal also comes after Mythic completed an oversubscribed $125 million funding round backed by investors including DCVC, NEA, Future Ventures, S3 Ventures, Atreides Management, and SoftBank.
Mythic Chief Executive Officer Taner Ozcelik said rising AI energy consumption is creating sustainability and infrastructure challenges across the industry.
He said the addition of Videantis strengthens Mythic’s ability to develop AI chips with significantly lower power requirements while also diversifying manufacturing capabilities across the United States and Europe.
Videantis, headquartered in Germany, is regarded as one of Europe’s established processor IP companies. The company was recently ranked among the top 1% of companies evaluated by the European Innovation Council and received one of the region’s competitive innovation grants.
According to the companies, Videantis processor IP has been deployed in more than 25 million chips globally, including automotive-grade systems used by major European vehicle manufacturers.
The company’s processor architecture supports workloads including deep learning, computer vision, image processing, video encoding and decoding, signal processing, and AI inference within a unified processing environment.
Unlike traditional GPU-centric systems that combine multiple specialized compute blocks, Videantis uses a scalable array of identical processing cores built around Very Long Instruction Word (VLIW) and SIMD architectures.
The companies said this design improves performance-per-watt efficiency while reducing the need for complex memory coherence systems and high-bandwidth interconnects.
Hans-Joachim Stolberg, founder and CEO of Videantis, said both companies share a common philosophy focused on minimizing data movement to improve energy efficiency.
He said combining Mythic’s analog compute-in-memory technology with Videantis’ digital architecture creates a more efficient hybrid AI computing system.
As part of the transaction, all five Videantis founders will join Mythic. Videantis will continue operating as a wholly owned subsidiary.
Mythic said the combined platform is designed to integrate analog computing for matrix operations with digital processing for control, programmability, attention calculations, SLAM, video processing, and other advanced workloads.
Dave Fick, co-founder and CTO of Mythic, said the company has long envisioned a hybrid computing model that combines analog efficiency with digital flexibility.
The acquisition is also expected to accelerate Mythic’s expansion into AI data centers, where growing power consumption has become a major operational challenge.
The company said its hybrid architecture is intended to improve AI performance per watt, per rack, and per dollar compared to existing GPU and TPU-based infrastructure.
Mythic added that the transaction establishes a transatlantic AI semiconductor platform combining Silicon Valley engineering with German semiconductor expertise as competition intensifies in the global AI infrastructure market.
