When you purchase through links on our site, we may earn an affiliate commission. This doesn’t affect our editorial independence.

Qualcomm has made a bold move toward creating the generative AI (genAI) capabilities of the future by purchasing the Vietnamese startup MovianAI. MovianAI is the generative AI research division of the Vietnamese IT giant Vingroup. The US company plans to leverage this acquisition to target a variety of devices, including smartphones.

The acquisition of MovianAI would enable Qualcomm to pursue “fundamental AI research,” the company stated. It intends to incorporate MovianAI’s products into phones, PCs, SDVs (Software-Defined Vehicles), and other devices.

According to a statement from the company, following the acquisition, Dr. Hung Bui, the founder and CEO of VinAI, was also hired by Qualcomm. Dr. Hung was a part of Google’s Deepmind project before founding VinAI. He talks about his time working with Google DeepMind, a US corporation, where he honed his skills in AI. He is renowned for his passion for pushing the boundaries of what is possible with technology.

Vietnamese online publication VNExpress claims that MovianAI is valued at about US$9 million. The precise terms and financial stakes of the acquisition were not disclosed by Qualcomm. Qualcomm described MovianAI as a company renowned for its proficiency in genAI, machine learning, computer vision, and natural language processing. Qualcomm thinks that MovianAI will enable them to produce “extraordinary inventions” in the field of genAI.

VinAI: History of Innovation and Excellence

Researchers at VinAI introduced a new method for creating image-generation AI models around two years ago. The method entails training a simpler algorithm that can accomplish the same tasks with less hardware utilizing an existing image-generating model. VinAI claims that the technique facilitates AI development by lowering the computing requirement for training models.

The lab has created a number of open-source AI models in addition to new machine-learning techniques. The lab combined three distinct neural network architectures, the Transformer, Diffusion, and Mamba architectures in one of its image-generating techniques. Mamba is a relatively new substitute for the transformer architecture that may be more efficient.

The research being done by VinAI goes beyond just image generation. It developed RecGPT-7B, a large language model that serves as recommendation generation software.

There appears to be a wave of acquisitions by big tech companies. The semiconductor giant, Nvidia was reportedly in advanced talks to buy Lepton AI. Lepton AI is a company that specializes in renting out servers powered by Nvidia’s AI chips. This would be the second acquisition after Nvidia recently acquired Gretel, a synthetic data business based in San Diego.

Check Out Our Previous Post:

Microsoft in Talks to Acquire TikTok, Says Trump

LEAVE A REPLY

Please enter your comment!
Please enter your name here