In brief: OpenAI has completed its first in-house developed chip, Jalapeno, designed for its AI models to directly address hardware requirements and increase efficiency.
OpenAI has unveiled its first chip, specifically developed for running its own AI models. In its development, the company collaborated with its models themselves as well as with an established semiconductor manufacturer.
OpenAI presents Jalapeno, its first proprietary chip, optimized for deploying the company’s own large language models. The in-house development is intended to directly address the hardware requirements of these models and thereby enable efficiency gains.
In addition to OpenAI, the company’s own AI models were involved in validation and optimization – an approach that allows for iterative improvements in hardware-software integration. Additionally, OpenAI cooperated with an established chip service provider to handle manufacturing and the complexity of chip development.
For CTOs, this signals a strategic shift: large model providers are diversifying their infrastructure and thus reducing their dependence on standard chips. This step points to increasing specialization in the AI hardware market, where the boundaries between software and hardware optimization are blurring.
Source: www.golem.de · Published 25 June 2026
Lumi AI News — AI-assisted curation in accordance with Article 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.7.1.