Sources inform Axios that Groq is aiming to secure $650 million in new financing from current investors, as it focuses on its inference neocloud sector, which depends on its proprietary AI chip and systems.

In December, Groq entered into one of those not-an-acquisition agreements with Nvidia for an estimated $20 billion, which included the exit of some senior Groq executives to the chip giant and the licensing of Groq’s hardware technology to Nvidia. This arrangement was a win for the startup’s investors, who received cash payouts from what would have represented Nvidia’s largest acquisition had the deal been a full buyout, Axios notes.

Currently, these investors have been asked to provide support for the company’s strategy to expand its inference cloud business, allowing developers and enterprises to run their inference-intensive applications. Inference refers to the processing that occurs following an AI prompt and is presently a more significant demand in the AI landscape compared to model training.

The new strategic direction is currently being guided by Groq’s interim CEO and CFO, Adam Winter and Matt Eng, respectively. 

In certain respects, the $650 million in funding appears to be assured. Axios reports that Groq’s investors Disruptive and Infinitium have consented to cover the round if other existing investors decide against their pro-rata allocations.