[ad_1]
Image: Microsoft
Off the back of making its Emissions Impact Dashboard (EID) tool widely available for Azure last October, Microsoft has announced the preview of EID for Microsoft 365.
In a blog post, Microsoft said by extending the tool it will help organisations quantify the greenhouse gas emissions associated with their usage of Microsoft 365 applications, including Exchange Online, SharePoint, OneDrive for Business, and Microsoft Teams.
Microsoft added that using the dashboard, organisations can measure total emissions and break that down to emissions per active Microsoft 365 user.
Customers can also segment and filter emissions by scope and data centre region, and estimate the carbon emissions they avoid by moving their Exchange and SharePoint deployments to the cloud.
“This data is crucial for understanding how usage of cloud services impacts emissions and is the first step in establishing a foundation to drive further decarbonization efforts,” Microsoft said.
According to Microsoft, the preview of EID for Microsoft 365 is a precursor for the upcoming general availability launch of Microsoft Cloud for Sustainability, a SaaS solution that organisations can use to record, report, and reduce emissions across their enterprise.
“Down the road, we plan to make the insights produced by the Emissions Impact Dashboard available via Microsoft Cloud for Sustainability. This will help customers gain a broader understanding of their environmental impact, end-to-end,” Microsoft said.
Introducing sustainability focused tools for Microsoft is part of the tech giant’s ambitious goal to be carbon negative by 2030, and that by 2050, it will have removed from the environment all the carbon that has been emitted by its business — directly and by electrical consumption — since it was founded in 1975.
Meanwhile, a study undertaken by Google Research has highlighted that accurate data is needed to help assess the carbon emissions in machine learning and identify strategies to mitigate those emissions.
The study, The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink, published in IEEE Computer identified there are four best practices that it claims can reduce energy by 100 times and carbon emissions by 1,000 times.
Known as the 4Ms, these best practices are: Model, machine, mechanisation, and map optimisation.
Under “model”, the study advises selecting ML model architectures, such as sparse models, while machine is about using processors and systems optimized for ML training, versus general-purpose processors.
The study also advises computing in the cloud rather than on-premise and choosing cloud locations with the cleanest energy.
Google said it is already using the 4Ms today and are available for users on its Google Cloud services.
“We demonstrate four key practices that reduce the carbon (and energy) footprint of ML workloads by large margins, which we have employed to help keep ML under 15% of Google’s total energy use,” Google Research brain team distinguished engineer David Patterson wrote in a post.
Like Microsoft, Google has also set its own green target and aims to run using carbon-free energy by 2030. The company has been carbon neutral since 2007, after it eliminated its entire carbon legacy.
Related Coverage
[ad_2]
Source link