Google slashes prices for its machine learning service as AWS steps up competition
摘要：Google has massively cut prices for its managed cloud machine learning service just two weeks after AWS released a competing offering at its re:Invent user conference.
Google has massively cut prices for its managed cloud machine learning service just two weeks after AWS released a competing offering at its re:Invent user conference.
The company has introduced massive price reductions for its Cloud Machine Learning Engine managed services. For example, customers using basic-tier compute for training a machine learning system will pay 43 percent less than they did earlier this year. Google also offered customers more clarity on what they’ll be paying for those jobs.
Information of the price reductions was first included in a blog post that appeared briefly yesterday on Google’s website, then vanished. A representative for the company declined to comment further on the news when reached for comment.
The price reductions come just weeks after Amazon Web Services introduced SageMaker, a service that’s supposed to help developers train and run custom machine learning models. Google’s ML Engine service offers similar functionality, including automatic tuning of algorithm parameters, along with a wealth of prebuilt machine learning systems.
Price cuts are to be expected from Google, which puts a great deal of work into cutting the costs of its offerings for customers in an already discount-happy business. These changes also bring some clarity to the messy pricing that has been ML Engine’s hallmark: Google is providing per-hour pricing for all of the different types of training machines available, something that was previously far harder to access.
Machine learning operations are a natural fit for cloud workloads. Deploying intelligent systems often requires a great deal of computing capacity and data storage, something that platforms like Google Cloud and AWS are well suited to provide. The perception that Google holds a lead in machine learning has helped the company secure high-profile customers, and these changes may make it more appealing than competing offerings.
For customers who want to run machine learning jobs on Google’s bare infrastructure, the company recently cut the price of attaching Nvidia Tesla GPUs to infrastructure-as-a-service instances available through its Compute Engine.
In addition to the price cuts, the leaked blog said Google plans to make its online prediction feature generally available, which would provide customers with an environment to request real-time predictions of machine learning data. That’s important for applications that must react as quickly as possible with intelligent predictions.
The company also plans to introduce a private alpha test of online prediction for gradient boosted decision trees made with the popular scikit-learn and XGBoost frameworks. That combination is frequently used to win machine learning competitions on Kaggle, a Google Cloud-owned site that lets data scientists tackle a wide variety of challenges to see who can discover the best approach.
Because the blog post outlining those changes has been pulled, it’s possible that the feature updates may change when Google opts to launch them at a later date. The pricing changes have already been made available through the cloud provider’s official documentation.