Google’s AI Chip Strategy Gets a Powerful Upgrade

Google is making another move towards competing against Nvidia in the AI space. The technology company has come up with two customized versions of the Tensor Processing Units (TPUs) developed in-house to address the increasing needs of firms and AI engineers.

The TPU 8tx and the TPU 8i Max are reworked versions of current processors and not new processors altogether. They each cater to a specific problem in the AI industry.

Security is the focus of TPU 8tx. The product is meant for government organizations as well as companies in certain industries that deal with confidential information. The system integrates Google’s chip technology with AMD processors and hardware-based encryption. The company states that it can provide double the speed performance compared to the previous model and still have good data security.

The TPU 8i Max is somewhat different. It is targeted at artificial intelligence systems that need to work with large amounts of data. The system uses two chips in one unit, which gives more power and memory. Therefore, it can be used for coding assistance, mathematical reasoning and drug discovery.

Google’s Move for the AI Market

It looks like Google has decided to monetize its proprietary chip technology into a proper business. The TPUs have been used in products like Search, YouTube and Gemini. This time, the company hopes that more enterprise customers will use Google Cloud to buy its hardware.

We believe that such an approach might totally revolutionize the AI infrastructure industry. The growing competition might assist companies in receiving better performance, greater security, and reduced costs for computer capacity. Moreover, the companies' dependence on Nvidia might be reduced.
And for end-users, it will be good for the development of AI technologies in the long term.

Are you interested in the most recent advances in technology? Share this article, subscribe to us on social media and read more about AI on our blog.