Recently, the U.S. artificial intelligence startup OpenAI issued a statement denying media reports about its plans to use Google's self-developed chips. This statement was released after related news from two days ago sparked widespread attention. A spokesperson for OpenAI said on Sunday that although the company is conducting early tests on Google's Tensor Processing Units (TPUs), it currently has no intention of using these chips on a large scale.
Image source note: The image is AI-generated, and the image licensing service provider is Midjourney.
In the field of artificial intelligence, it is common for laboratories to test various chips, but achieving large-scale application of new hardware usually takes a long time. In addition, this involves different architectures and software support, which is quite challenging. OpenAI stated that it is currently actively using graphics processing units (GPUs) from NVIDIA and artificial intelligence chips from AMD to meet the growing computing demands.
Additionally, OpenAI is also developing its own chips, which are expected to reach a "mature" milestone this year, at which point the design of these chips will be finalized and put into production. Previously, there were reports that OpenAI had reached a cooperation agreement with Google Cloud to meet its increasing computing needs. This collaboration was seen as an unexpected partnership between two major competitors in the artificial intelligence field.
Although OpenAI's main source of computing power will be GPU servers supported by the emerging cloud company CoreWeave, Google is also actively expanding the external availability of its self-developed artificial intelligence chips (TPUs). TPU chips were previously used mainly for internal projects at Google, but they are now beginning to attract the attention of other tech giants such as Apple, as well as some startups like Anthropic and Safe Superintelligence, which are competitors of OpenAI.
OpenAI's latest statement clearly conveyed its position on the cooperation rumors, emphasizing the company's diversity in hardware usage and its determination to develop its own technology. With the rapid development of artificial intelligence technology, OpenAI is clearly facing challenges and opportunities from all sides.
Key points:
🌐 OpenAI denied the plan to cooperate with Google's self-developed chips, stating that it is not currently deploying them on a large scale.
🖥️ OpenAI is currently using chips from NVIDIA and AMD, and is developing its own chips to meet computing demands.
☁️ Although cooperation with Google Cloud was reported, OpenAI's main computing power comes from GPU servers provided by CoreWeave.