OpenAI announced Tuesday that testing results of its new custom inference chip, called Jalapeño and developed in partnership with Broadcom, show a significant performance advance. The chip can process more AI workloads per unit of power while delivering faster responses. OpenAI plans to begin deploying Jalapeño in its compute infrastructure by the end of the year, which should support Broadcom’s FY2027 AI revenue outlook exceeding $100 billion. However, this news does not mark the end of OpenAI’s partnership with Nvidia, as Nvidia recently committed up to $105 billion in credit support for OpenAI’s data center project in Ohio, which will exclusively use Nvidia compute.
Source: Read the original article
