AI summary
Summary of article:
People familiar with the matter said OpenAI may buy AI chips from Google to expand its computing capabilities; At the same time, its self-developed chip plan and Oracle’s arrangement to provide TPU resources are also making progress.
People familiar with the matter disclosedOpenAI IntentionGoogle The procurement of artificial intelligence (AI) computing chips as part of a broader layout to expand AI infrastructure.
Bloomberg News reported that the final terms of any deal between OpenAI and Google are still unknown, but it shows that OpenAI is taking a multi-pronged approach to expanding AI computing capabilities. Both OpenAI and Google declined to comment.
OpenAI has previously announced that it plans to start using its own chips to support AI models later this year. Richard Ho, who leads the internal chip project at OpenAI, once participated in the development of Google’s tensor processor (TPU). After joining OpenAI, they worked togetherbroadcom (Broadcom) jointly developed their own jalapeno chip.
Oracle also announced on the 27th that it will provide TPU computing resources in its cloud computing infrastructure.
Summary FAQ
-
The report notes that OpenAI’s move aims to expand AI infrastructure and computing capabilities and is part of a broader layout. This suggests that the company hopes to obtain chip resources from multiple sources to support model training and service needs.
-
OpenAI has previously announced that it plans to use its own chips to support AI models later this year; Richard Ho, who is responsible for internal chip planning, has participated in the development of Google TPU and jointly developed the Jalapeno chip with Broadcom.
-
In addition to the reported intention to purchase OpenAI, Oracle also announced on the 27th that it will provide TPU computing resources for its cloud computing infrastructure, indicating that Google TPU is expanding market applications through various cooperation models.
