OpenAI Plans AI Chip Deployment by Year-End

According to media reports, OpenAI plans to begin deploying its internally developed AI chip, Jalapeño, in collaboration with Broadcom by the end of this year. The company is also working with NVIDIA, AMD, and other partners to advance the development and deployment of AI accelerators.

The reported chip initiative comes as OpenAI continues to expand its AI computing infrastructure and reduce its reliance on external hardware. Collaboration with major semiconductor companies could further strengthen its access to high-performance AI acceleration technologies.

Separately, OpenAI is reportedly preparing to launch a powerful new AI model called Astra. The company is expected to initially restrict access to some of the model’s most advanced cybersecurity capabilities.

In a blog post published Tuesday, OpenAI said the new AI model would be released “soon,” but did not provide a specific launch date. Initially, only a limited group of testers will be able to use the model for advanced cybersecurity-related tasks.

OpenAI plans to gradually expand access to defensive cybersecurity applications through its Daybreak Blue program. Under the initiative, approved testers can use the company’s most capable AI models with additional safeguards designed specifically for cybersecurity-related work.

The developments highlight OpenAI’s efforts to strengthen both its AI computing infrastructure and cybersecurity capabilities, while its partnerships with Broadcom, NVIDIA, AMD, and other semiconductor companies could play an important role in supporting future AI workloads.