Amazon Web Services boosts open source credentials
Amazon has taken another step that reaffirms its commitment to contribute to open source; a commitment critics have long dismissed as "not enough", "window dressing" and "insufficient" given its own consumption of and dependence on open source.
Last week, Amazon's cloud computing unit, Amazon Web Services (AWS), the world's largest cloud vendor, announced it was releasing the code of Amazon SageMaker Neo as the open source Neo-AI project under the Apache Software Licence.
This follows last week's other big open source announcement from Microsoft, which revealed its intention to add San Francisco-based Citus Data to its growing open source portfolio alongside the world's largest repository of open source code, GitHub.
Like AWS, Microsoft's open source commitment and bona fides are regularly subjected to scepticism or outright criticism from industry analysts and open source purists. This is despite the fact the two companies are among the top 10 contributors to open source projects held in the GitHub repository, according to research published last year by WhiteSource, a licensing, security and reporting solution for managing open source components.
With 4 550 contributors, Microsoft far outstripped its closest rival, Google, in the contribution stakes and, in total contrast to its positioning as an "anti-open source company", is now what WhiteSource maintains is "the biggest advocator for open source".
AWS was placed sixth on the WhiteSource list, behind Microsoft, Google, Red Hat, IBM and Intel. With 881 contributing employees, it provides "great projects on GitHub for use by the open source community for working with their different products like Alexa", WhiteSource noted.
When AWS originally launched SageMaker Neo, a new machine learning feature that can be used to train a machine learning model once and then run across multiple environments, at re:Invent in November 2018, there was no mention of its coming open source status.
However, it was noted at the time that SageMaker Neo was itself derived, in part, from other open source projects, including Treelite and TVM.
According to Sukwon Kim, senior product manager for AWS Deep Learning, and Vin Sharma, engineering leader for AWS Deep Learning, prior to the release of SageMaker Neo, it was difficult to optimise a machine learning model for multiple hardware platforms because developers had to tune models manually, with considerable trial and error, for each platform.
Now, with the release of AWS code back to open source through the Neo-AI project, any developer will be able to innovate on the production-grade Neo compiler and runtime.
"By working with the Neo-AI project, processor vendors can quickly integrate their custom code into the compiler at the point at which it has the greatest effect on improving model performance. It also enables device makers to customise the Neo-AI runtime for the particular software and hardware configuration of their devices.
"The Neo-AI project will absorb innovations from diverse sources into a common compiler and runtime for machine learning to deliver the best available performance," they stated.