In November 2015, tech giant Google announced that it is releasing its in-house machine learning software TensorFlow in the open-source platform. Recently, Google upgraded TensorFlow by releasing a distributed version of the program that allows it to run in multiple machines, which Google claims could reach hundreds at a time.

Like Us on Facebook

Several tech analysts agree that Google's latest upgrade to TensorFlow is the most logical step since the machine learning software gains more power as it analyzes more and more amounts of data. By feeding TensorFlow huge amounts of data, the software gets to create and develop trends like facial features in photographs.

Despite loads of positive reactions given to the latest TensorFlow update, the team behind the project claim that they actually faced a couple of delays in releasing the multi-server version of the program.

In a statement, TensorFlow technical lead Rajat Monga said, "Our software stack is differently internally from what people externally use. It would have been extremely difficult to just take that and make it open source."

By allowing TensorFlow to run several operations on vast networks of computers simultaneously, the program can make smarter systems faster.

Along with the new TensorFlow update, Google also highlighted the success of the program. Google claims that TensorFlow is the most popular piece of machine learning framework on GitHub. Additionally, it is also the most forked project on the site in 2015 despite the fact that it was only released in November last year.

Since its release, TensorFlow has been used by several developers on a number of different yet equally interesting projects. One of these projects is a program that learns, through trial-and-error, how to play "Pong." Another project is a neural network that can invent fake yet realistic looking Chinese characters.

TensorFlow operates on a branch of artificial intelligence called deep learning. Deep learning is used by other tech giants like Microsoft and Facebook in order to power their own artificial intelligence projects.

TagsGoogle, TensorFlow, Google TensorFlow, TensorFlow update, TensorFlow new update, Google update, Google TensorFlow new update