Hark signs a multi-year deal with NVIDIA for gigawatt-scale compute on Vera Rubin
The new company from Archer and Figure founder Brett Adcock is building a multimodal assistant that remembers its user. The agreement covers both model training and the inference behind its launch.
AI startup Hark has announced a multi-year strategic agreement with NVIDIA. At its centre is gigawatt-scale compute capacity on NVIDIA's next-generation Vera Rubin platform.
Hark builds multimodal systems that work through speech and vision and hold a persistent memory of who the user is and what they are trying to do. The company is developing both the software and its own hardware, aiming to deliver that interface through existing devices as well as equipment it makes itself. Its founder and chief executive is Brett Adcock, who previously founded Archer Aviation and Figure.
The announcement lists what the deal covers: training the company's own foundation models, the distributed inference that will run the platform at launch, and Hark Handoff, its computer use model. On the NVIDIA side, the work draws on Megatron, Dynamo, Nemotron pretraining data and NVSentinel.
No financial figure appears in the announcement, and the megawatt breakdown of the capacity is not disclosed. NVIDIA was represented by Nico Caprez, Vice President of Global AI Infrastructure Growth.
The full announcement is on Hark's site.
