Openai recently began renting Google's artificial intelligence chips to power ChatGpt and other products.
CHATGPT makers are one of the biggest buyers of NVIDIA's Graphic Processing Units (GPUs) using the process of using AI chips to train models, and using AI models for inference computing, and using the AI models to make predictions or decisions based on new information using trained knowledge.
Openai had planned to add Google Cloud Service to meet its computing power needs. Reuters reported exclusively earlier this month, showing an astonishing collaboration between two prominent competitors in the AI sector.
In the case of Google, transactions are made as it expands the external availability of internal tensor processing units (TPUs) that were historically reserved for internal use. This helped Google attract customers, including startups such as Big Tech Player Apple, two ChatGpt-Maker competitors launched by former Openai leaders, and startups such as artificial and secure super intelligence.
Google's move to rent TPUS signals shows Openai is using non-Nvidia chips significantly for the first time, leaving Sam Altman-led company away from relying on backer Microsoft's data centers. According to information previously reported development, it could boost TPU as a cheap alternative to Nvidia's GPUs.
Openai hopes that TPUs renting through Google Cloud will help reduce the cost of inference, according to reports.
However, Google, a competitor of AI race Openai, cites Google Cloud employees, saying it doesn't rent the most powerful TPUs from its rivals.
Google declined to comment, but Openai did not respond immediately to Reuters when contacted.
Google's addition of OpenAI to its customer list shows how the tech giant has leveraged its in-house AI technology, from hardware to software, to accelerate its cloud business growth.
