OpenAI recently collaborated with AI infrastructure company Cerebras to add an additional 750 megawatts of low‑latency AI compute to its platform. This agreement will make the ChatGPT service respond faster to user inputs, according to the companies.
A staged deployment of the additional capacity is due to begin this year and conclude in 2028.
Budding Nvidia rival Cerebras states its technology accelerates AI inference by combining massive compute, memory and bandwidth on a single large chip, and removing bottlenecks found in traditional hardware.
Sachin Katti, head of compute infrastructure at OpenAI, stated Cerebras’ “dedicated low-latency inference” capabilities would deliver “faster responses, more natural interactions and a stronger foundation to scale real-time AI to many more people”.
Cerebras CEO and co-founder Andrew Feldman said “real-time inference will transform AI, enabling entirely new ways to build and interact” with models.
Bloomberg reported the deal is worth more than $10 billion.
OpenAI is pouring large amounts of money into data centres and compute infrastructure to compete with Meta Platforms, Microsoft and Google.
Source: Mobile World Live
Image Credit: OpenAI
