Nvidia announced on Monday that its Groq 3 LPX rack is now in full production, marking the first commercialization of technology from the company’s largest acquisition to date.
This new rack will be deployed alongside Vera central processors and Rubin graphics processors at Neocloud Nebius and is scheduled to go online later this year, according to Nvidia senior director Dion Harris.
The company’s efforts to manufacture and deliver Groq’s chip emphasize the increasing importance of low-latency inference, which is crucial for making AI agents feel more responsive and reducing delays for users, particularly in coding applications.
Nvidia notes that cloud providers can charge a premium for these fast, efficient tokens.
Source (CNBC)
