Get the latest on Google’s self-designed Tensor Chips. The powerful chips that will power its next-generation Artificial Intelligence. Learn about the power and potential of these new chips and what they mean for the future of AI.

Machine learning (ML) revolutionizes technology every few years, and Google products have played a role in this advancement. Google Assistant has made devices more useful, and Google Translate has helped break down language barriers. However, we have not always been able to bring the full potential of ML to smartphones.

Generation of AI and Machine Learning Technologies

Google’s self-designed tensor chips, called Tensor Processing Units (TPUs), are already powering the company’s artificial intelligence and machine learning technologies. These chips are designed to accelerate the processing of deep neural networks, which are used in many applications, including image and speech recognition, natural language processing, and autonomous driving.

Google’s first generation of TPUs was released in 2016, and since then, the company has continued to improve the technology. In 2018, Google released the second generation of TPUs, which are even more powerful than the first. These chips are now being used in Google’s data centers to power its cloud services, including Google Cloud AI Platform and Google Cloud Machine Learning Engine.

In addition to the second-generation TPUs, Google has also announced the third generation, which it calls Cloud TPU v3. These chips are even more powerful than the second generation and are designed to be used for both training and inference in machine learning models. Google claims that a single Cloud TPU v3 can deliver up to 420 teraflops of performance.

Google’s investment in developing its own tensor chips is a testament to the importance of AI and machine learning for the company’s future. By designing its own chips, Google can optimize its hardware and software to work seamlessly together. Resulting in faster and more efficient processing of machine learning models. This will enable Google to continue to push the boundaries of what’s possible in AI and machine learning. Create new applications and services that were previously not feasible.

Tensor Chips Built with Google Research

Several years back, a group of Google researchers collaborated in the areas of hardware, software, and ML to create the ultimate mobile ML computer. This was done to achieve our vision of what we believed was possible on our Pixel smartphones.

Google Tensor enables us to create advanced machine learning features like Motion Mode, Face Unblur, Speech Enhancement Mode, and HDRnet for videos. By utilizing Google Tensor, we can enhance the usefulness of smartphones. Making them more intelligent and better suited to accommodate our individual needs.