Global tech behemoth, Google LCC has recently joined forces with social media giant, Facebook, Inc. to develop individual artificial intelligence (AI) technologies, further making them more compatible and efficient.
As per an official statement, Google and Facebook claim that a number of engineers are currently collaborating to make Google’s custom computer chips for machine learning application, dubbed Tensor Processing Units (TPU) compatible with Facebook’s open source machine learning PyTorch framework.
Rajen Sheth, Google Cloud’s Director of Product Management said that Google’s TPU team engineers are actively working with core PyTorch developers to link PyTorch to Cloud TPUs as to enable users to enjoy the flexibility and simplicity of PyTorch, while benefiting from the scalability, performance, and cost-efficiency of Cloud TPUs.
For the record, in 2016, Google first announced its TPUs during its annual developer conference, claiming it to be a more efficient way for researchers and companies to power their machine-learning software projects.
Sources cite that Google sells access to its TPUs through its cloud computing business instead of selling the chips separately to customers like Nvidia, renowned for its graphics processing units (GPUs) among researchers currently working on deep learning projects.
Incidentally, AI technologies like deep learning have grown in popularity over the years with tech giants like Facebook and Google using the technologies to create software applications that can automatically perform tasks like identifying images in photos.
As more businesses move towards machine learning technology, companies like Facebook, Google, and others have now created their own Artificial Intelligence software platform, specifically coding tools, intended to make it easier for developers to create their own machine-learning powered software.
Reliable sources claim that data scientists and machine learning engineers now have the availability of a wide range of open source tools for developing intelligent systems and the recent collaboration would be a critical step towards providing users access to high-end hardware and software capabilities via AI models.