Google to equip data centre infrastructure with AMD deep learning hardware.
The game of one up-manship in the cloud market seems never ending, which is why it should come as no surprise that on the day Microsoft signed up a big name artificial intelligence company, Google would reveal some major developments in the same field.
Firstly there is the development that Google is to equip its data centre infrastructure with AMD’s graphics processing units (GPUs) in order to accelerate deep learning applications.
The move is AMD’s first real push into the deep learning market while one of its main rivals, Nvidia, has been making strides in this area for some time.
Google will start to roll out AMD’s FirePro S9300 x2 GPUs in 2017 and will be applied to the Google Compute Engine and Google Cloud Machine Learning Service. Other GPUs available will be the Nvidia Tesla P100 and Nvidia Tesla K80s.
The AI and machine learning news from Google didn’t stop there as it revealed on its blog that it has new offerings in Cloud Machine Learning.
To support the company’s efforts in developing machine learning capabilities it has created the Google Cloud Machine Learning group. This group will be focused exclusively on delivering cloud-based machine learning solutions to all businesses. The idea is to build a centralised team within Google Cloud so that it can accelerate its ability to deliver machine learning products and services.
The company said that this: “marks an exciting next step in Google Cloud’s product commitment to making machine learning more accessible for all businesses.”
To highlight the importance of machine learning to the company it decided to introduce a new Cloud Machine Learning API that’s designed to help people find careers, new hardware options, improved efficiencies and expanded features for Cloud Translation, Cloud Vision and Cloud Natural Language APIs.
The Jobs API is designed to provide businesses with: “Google-strength capabilities” that will help them find, match and recommend relevant jobs to candidates. It uses machine learning to understand how job titles and skills relate to one another and what job content, location, and seniority are the closest match to a jobseeker’s preferences.
Google also revealed that it has reduced the prices for Cloud Vision API by up to 80%.
The company didn’t stop with the machine learning announcements there as it also revealed Cloud Translation API Premium. The company is offering its Neural Machine Translation system in a premium edition of Cloud Translation API.
Google’s Cloud Natural Language API has also been moved to general availability with additional features such as granular sentiment analysis with expanded language support and improve syntax analysis with additional morphologies such as number, gender, person and tense.
It’s no surprise that Google has decided to go all-in with its machine learning capabilities and create a more centralised strategy. One of the company’s cloud strengths has long been in these kinds of fields and it certainly won’t want to be left behind by the likes of Microsoft.