Google Announces 2nd-Gen AI Chip | PYMNTS.com

This post was originally published on this site


Google just made a major move to expand its revenue beyond the realm of digital advertising.

At the keynote address of this year’s I/O developer conference, Google’s CEO announced that the company will be selling AI computer chips, called Cloud Tensor Processing Units (TPUs), via Google Cloud service.

Urs Hölzle, Google’s veteran technical chief, told Bloomberg that the chip is “basically a supercomputer for machine learning. The field is rapidly evolving. For us, it’s very important to advance machine learning for our own purposes and to be the best cloud.”

Machine learning is, in essence, a form of artificial intelligence that allows programs to change, adapt and (namely) learn from new data, without requiring additional programming effort and human intervention.

These intuitive computing capabilities look to innovate and automate repetitive, highly difficult pattern-matching problems and insight generation across massive volumes of data. For Google, machine learning powers text translation, search rankings and voice recognition.

Bloomberg noted that Google created the chip to address issues around the high cost and high demand on computing power machine learning took up in the company’s data centers. Hölzle was quoted as saying that a single Cloud TPU device, composed of four chips, is some 12,000 times faster than IBM’s Deep Blue supercomputer.

The news of the chip comes along with the announcement of machine learning innovations across Google’s products, reportedly including a new photo editing tool, features for Google Assistant and a new web portal for the company’s AI plays.

It’s all a part of Google’s push to grow its cloud business. The move looks to place Google in a position to fight for the top two spots in the cloud computing space currently held by Microsoft and Amazon Web Services.

Buyers will need to sign up for a Google cloud service, run their tasks and store their data on Google equipment, noted Bloomberg, in order to get the Cloud TPU chip.