Tech giant Microsoft recently announced that it is open-sourcing its deep learning toolkit. Just like Google, Microsoft decided to open source the Computational Network Toolkit which the company claims is more efficient compared to four other popular deep learning tools.

In a statement posted on the company's official blog, Microsoft said, "Over the past few years, the field of deep learning has exploded as more researchers have started running machine learning algorithms using deep neural networks, which are systems that are inspired by the biological processes of the human brain. Many researchers see deep learning as a very promising approach for making artificial intelligence better."

Like Us on Facebook

Microsoft's Computational Network Toolkit (CNTK) is available for those interested to use it and it can be downloaded via Github. Microsoft said that the toolkit is available for anyone including start-up companies dealing with deep learning technologies to large corporations that want to process huge amounts of data in real time.

Microsoft claims that CNTK is more efficient compared to four of its closest rivals; TensorFLow, Theano, Caffe and Torch 7. The project's chief scientist, Xuedong Huang said that the toolkit was developed out of necessity as they wanted to improve how current computers understand speech, but most of the current tools they use are slowing down their progress. Out of this necessity, the team decided to build their own.

At Microsoft, CNTK runs on an array of computers that users GPU's or graphics processing units. Although GPU's are designed to render graphics, they are also ideal for crunching huge algorithms which can lead to advances in deep learning and artificial intelligence.

Unlike its rivals, CNTK can scale across several GPU-based systems. This makes it ideal in creating large clusters of GPU-based computers which will be essential in major experiments as well as large calculations.

Microsoft claims that although deep learning is still in its infancy, the platform will have huge effects in the future of information technology. Currently, researchers are starting to run machine learning programs by employing deep neural networks which are designed to mimic the functions of the brain, according to PC Mag.