More than ever, leaders are under pressure to reduce their environmental impact. This is especially true for data centers due to their contribution to global warming. If all the data centers in the world were one country, it would be the 5th largest energy consumer in the world. In 2020, data centers consumed about 1% of the world’s electricity needs and contributed to 0.3% of all CO2 emissions.
Businesses are now required to be transparent about their carbon footprint, and data centers are in a race to improve their efficiency rankings. There is a list of data centers around the world that are rated by PUE (Price-Use-Effectiveness) and Greenpeace has produced a cleantech industry ranking of centers based on their carbon footprint.
The need for greener code
Many of the data center sustainability initiatives are based on using renewable energy for cooling or optimizing cooling systems to reduce power consumption. However, in addition to the energy required to maintain the environmental conditions required for data analysis, the software itself also has a significant impact on the amount of power consumed. How much? Pretty much.
Based on recent research, a large machine learning (ML) model like Meena uses the same amount of energy as a passenger car driven 242,231 miles. Researchers at the University of Massachusetts at Amherst estimated that training a large deep learning model produces 626,000 pounds of CO2, equivalent to the lifetime emissions of five cars.
As a result, there is an increased interest and dedication to creating more efficient code. The Green Software Foundation (GSF), with members such as VMware, Microsoft, Accenture and GitHub, has a mission to design, develop and code software that uses less energy.
Tips for sustainable machine learning
There are several scholarly articles on how to write greener algorithms for AI/ML models, but here are a few basic tips.
One way to reduce computational resources is to minimize the number of training experiments. There are hundreds of pre-trained ML models, or blueprints, where developers only need to infuse their own data to infuse AI capabilities into applications, greatly reducing the time it takes to develop and train models.
It is also important to have insight into the carbon footprint of the algorithm in order to make decisions about the best way to optimize performance. Researchers from several universities have developed tools for this purpose. For example, Green Algorithms calculates your cloud computing carbon footprint. Another example is CodeCarbon, a software package that integrates with the Python code base and estimates the amount of CO2 generated by the computing resources used to run the code.
Automation can also be used to reduce training run time. It is possible to minimize the number of experiments and/or the amount of data analyzed while maintaining accuracy. More efficient data sampling alone can speed up model runtime by a factor of 5.8.
The software used to actually perform the calculations can also help reduce the number of computing resources required. There are databases specifically designed to handle large amounts of data that can optimize the use of memory and storage to reduce energy consumption. These databases also have the advantage of not having to limit the amount of data to be analyzed, reducing the risk of compromising the accuracy of the model by trying to speed up runtime.
The reduction in model runtime, in addition to increasing energy efficiency, reduces the total time to insights for business-critical applications such as fraud detection, cybersecurity solutions, quality control, etc. More efficient code is not only better for the environment, but also good for business.
More and more potential customers want transparency about a company’s commitment to its green strategies, and adopting a ‘green’ code standard could be an important first step. Employees want to work for an ecologically sensitive company that makes responsible decisions regarding the environment. In the future, cloud providers could require visibility into a workload’s carbon footprint, with fines for processing deemed excessive or unnecessary.
With the sheer number of calculations required to derive the meaning of better business decisions, social responsibility has become not just a nice touch, but a necessity.
Ohad Shalev is a Strategic Analyst at SQream.
data decision maker
Welcome to the VentureBeat community!
DataDecisionMakers is the place where experts, including technical staff, working with data can share data-related insights and innovations.
If you want to read about innovative ideas and up-to-date information, best practices and the future of data and data technology, visit us at DataDecisionMakers.
You might even consider contributing an article of your own!
Read more from DataDecisionMakers