Google, a subsidiary of Alphabet, is launching more types of AI chips in an attempt to quickly catch up with Nvidia in this hundreds of billion dollar market. Google has launched two new versions of tensor processing units, which are custom chips for training and running AI software. The new chip can meet the needs of confidential computing, allow the model to retrieve a wider range of information when processing questions, and adapt to the major trend of the industry shifting to AI agents while facing various security challenges. This release marks an acceleration in the iteration of Google's chip division. In the past, the business launched a new AI processor every two years, and released two at once this year; the 8tx and 8i Max variants, which will be officially announced on Monday, are new segmented versions added to these two chips. Armin Waddett, the head of the Google chip, said in an interview that each of the four TPU products is adapted to different needs. “This is in line with Google's core judgment: the future will move towards specialization, and the profit margin brought by specialization is still huge.”

Disclaimer:Webull uses external vendor Google Translation Service for news translations where we endeavour to ensure these are correct, however, we recommend that you please double-check this information accordingly. Webull is not responsible for translation errors or issues.