Google Launches TensorFlow Enterprise at No Extra Cost for Cloud Customers

Google has announced the launch of TensorFlow Enterprise, a version of its popular open source software library for machine learning projects — and it comes at no additional cost beyond that of the cloud platform on which it runs.

First developed by Google's Brain division for internal use then released to the public as an open source project in 2015, TensorFlow can be found at the heart of a range of machine learning projects from automated LEGO sorters to monitoring a 3D printer for failure states. The tool isn't just popular with hobbyists, though: It's increasingly found powering larger commercial endeavours, and it's for these users Google has announced TensorFlow Enterprise.

"TensorFlow Enterprise incorporates: Enterprise-grade support; cloud scale performance; managed services," writes Google's Craig Wiley, director of product management for the company's cloud AI platform. "Together, these services and products can accelerate your software development and ensure the reliability of your AI applications."

Possibly the biggest feature of TensorFlow Enterprise is the promise of long-term support - up to three years, for certain releases. "The pace of AI and software versions is evolving rapidly, but many customers have told us they are heavily invested in a previous version of TensorFlow," Wiley explains. "That is why TensorFlow Enterprise includes long-term version support. For certain versions of TensorFlow, we will provide security patches and select bug fixes for up to three years. These versions will be supported on Google Cloud, and all patches and bug fixes will be available in the mainline TensorFlow code repository."

TensorFlow Enterprise also comes with a "white-glove service" support offering, which includes engineer-to-engineer support, and the promise that TensorFlow Enterprise ties in to other Google Cloud managed services such as the AI Platform and Kubernetes Engine.

Google has already signed up some big-name customers to the platform, including 3D engine creator Unity and automotive specialist GM Cruise, and has released a technical blog post for those looking to find out more. For the hands-on learner, there's a codelab featuring an example project for fraud detection.

TensorFlow Enterprise comes at no cost, other than those over the underlying cloud platform; the "white-glove service," meanwhile, is restricted to "select applicants."