Sources: Naveen Rao’s new AI hardware startup targets $5B valuation with backing from a16z 

Naveen Rao, who was previously the head of artificial intelligence at Databricks, is in conversation to pick up $ 1 billion on a $ 5 billion appreciation for a new startup, called UNCendental, Inc., which is building a new type of computer, according to four people who are familiar with discussions.

Andreessen Horowitz has agreed to lead the investment, and Lightspeed and Lux ​​Capital also participate in the deal, said two of the sources, although none of those VC companies responded to our request for comment. Databricks would also invest in the new RAO company, Bloomberg reported last month.

RAO has already picked up hundreds of millions and is planning to build the startup without waiting for the full round of $ 1 billion to close, according to our sources. He is planning to increase the rest of the capital in installments, a financing approach that is usually called a “managed” round.

RAO refused to comment, although he publicly New startup on X Last week he confirmed the name and describing the hoped product as “reconsidering the base of a computer to build a new substrate for intelligence that is just as efficient as biology. Efficiency of the brain scale without the biological baggage!”

Databricks took over RAO’s previous startup, Mosaicml in 2023, for $ 1.3 billion. Founded by RAO in 2020 with a focus on training and the use of large AI models, Mosaicml raised $ 33.7 million from investors, including Lux Capital, DCVC, Playground Global and Samsung, according to PitchBook. Prior to Mozaicml, RAO was co-founder of the Machine Learning Platform Nervana Systems, that Intel Corp. In 2016 took over for Reportedly More than $ 400 million.

Rao, who spent more than two years as VP of AI at Databricks (now valued at $ 100 billion and $ 4 billion in ARR), left the company last month to concentrate on his new startup, according to his LinkedIn profile and earlier reporting by Bloomberg.

Our sources say that Rao’s vision will essentially compete with Nvidia by designing a new AI machine that includes both adapted silicon chips and server infrastructure.