Artificial intelligence and generative AI (GenAI) present transformative opportunities to tackle sustainability objectives. With each advancement, AI’s ability to solve complex environmental issues improves, sparking hope for a more sustainable future. However, it is crucial to strike a balance. Choosing the right data storage can advance both the effectiveness of AI in addressing sustainability challenges and ensure AI’s own sustainability.
Addressing Sustainability Through AI
The potential of AI and GenAI to accelerate organizations’ sustainability transformations and drive business innovation is immense. A recent IDC worldwide IT buyer survey found that more than three-quarters (76%) of IT decision makers consider AI and its derivatives to be “critical” or “very important” for organizations’ journey to sustainability.1 Further, more than 40% said that at least half of their AI spend has a sustainability component.2
Addressing sustainability through AI involves leveraging advanced technologies to analyze vast amounts of data from various sources, such as satellite imagery, sensors and historical records, to identify patterns, predict trends, optimize resource use, and preemptively address supply chain risks. AI can be used to reduce waste in manufacturing, design and manage more energy efficient buildings, manage water resources more efficiently, optimize packaging and enhance agricultural practices to minimize environmental impact.
The selection of the right data storage solution plays a pivotal role in facilitating the effective use of AI to address sustainability. A well-suited solution can facilitate the ingestion and processing of vast amounts of complex environmental data by providing high throughput and low latency, ensuring rapid data access and real-time processing capabilities. Additionally, by ensuring scalability and advanced data management features, the right storage allows AI systems to accommodate increasing volumes and variety of data, without compromising performance or incurring excessive expenses. Scalable storage also facilitates seamless access to data across teams and geographic locations, promoting both collaboration and innovation. Additionally, the proper storage maintains the integrity and security of environmental data, ensuring that AI models trained on this data produce reliable and trustworthy results.
For example, Nature Fresh Farms, one of North America’s largest vertically integrated greenhouse-grown produce leaders, uses machine learning (ML) and automation to grow and deliver more sustainable fresh food solutions. Nature Fresh Farms’ AI-enabled edge solutions collect and store an average of 11.3 MB of data per plant per month for two million plants, and the farm integrates 23 TB of data every year across hundreds and thousands of sensors in its facilities. And there is also transactional data from everything from climate control reads to video files.
At the edge, Dell PowerScale storage supports Nature Fresh Farms’ massive workloads for everything from the video systems and metrics to Amazon Simple Storage Service (S3) buckets, to all the unstructured data for general files. PowerScale delivers the data needed for high-performance ML and deep learning. With superior capabilities for low latency, high throughput and massively parallel I/O, it is the ideal storage complement to GPU accelerated compute for AI workloads, effectively compressing the time needed for training and testing analytical models for multi-petabyte data sets.
With the PowerScale solution, Nature Fresh Farms has reduced the upload time for produce images from six to 12 hours overnight to real-time during the day, with growers able to access all the analytics and insights they need in under an hour. Nature Fresh Farms has used these AI-driven insights to increase yield per acre by approximately ten times compared to traditional farming and used collected sensor readings for a closed-looped system to recycle 97% of water.
Ensuring AI’s Own Sustainability
As AI algorithms require vast computational resources for training and inference tasks, organizations must balance the trade-off between innovation and environmental responsibility. In the survey cited earlier, IDC found that organizations’ main sustainability concern when deploying or using AI/GenAI is reducing energy demand and carbon emissions (30%), followed closely by electronic waste (e-waste) and IT end-of-life management (29%).3
Reducing energy consumption and e-waste in AI involves a multifaceted approach. Optimizing algorithms’ energy efficiency can significantly decrease the power requirements. Utilizing energy-efficient hardware components, such as low-power processors and accelerators, can shrink carbon footprints. Utilizing systems engineered for improved cooling efficiency and optimizing data center designs for airflow management can further decrease energy usage. Extending the lifespan of AI hardware through modular design and upgradability can reduce the frequency of electronic waste generation. And implementing responsible recycling and disposal practices for outdated components is crucial to minimizing e-waste and maximizing resource recovery.
Building AI solutions with sustainable storage—like PowerScale, the world’s most efficient scale-out file storage solutions4—to minimize environmental impact, is vital. There are four key ways where Dell is advancing the sustainability of PowerScale: energy efficiency, thermals and cooling, infrastructure consolidation and retire and recycle.
-
- Energy efficiency. We are decreasing the energy intensity of our technology, enhancing its efficiency with each successive generation. PowerScale innovations like all-flash drives, high-speed in-line data compression and deduplication enables more efficient resource use. Flexible as-a-Service consumption, such as Dell APEX Data Storage Services File, can help you easily scale on demand, reducing overprovisioning and energy usage.
- Thermals and cooling. We are innovating thermals and cooling to reduce heat as we increase system processing capabilities, while improving performance. For example, the latest Dell PowerScale F210 and F710, with Smart Flow design, a feature within the Dell Smart Cooling suite, streamlines airflow, directing the right amount of air to where it is needed to deliver up to 90% greater performance per watt.
- Infrastructure consolidation. We are enabling you to streamline your infrastructure to minimize physical footprint and carbon emissions. PowerScale’s scalable architecture allows you to consolidate workloads onto fewer systems, leading to a smaller physical footprint. Dell CloudIQ uses a rich set of patented AI and related algorithms to track and forecast storage energy and carbon footprint for making better decisions about consolidating workloads.
- Retire and recycle. We are helping you responsibly retire IT equipment and maximize its reuse potential. PowerScale storage is engineered with eco-friendly materials and designed for longevity. Dell’s recovery and recycling services can help you responsibly retire IT equipment and maximize its reuse potential to minimize e-waste.
In summary, AI offers transformative potential in tackling environmental challenges and inspiring optimism for a greener tomorrow. However, like any technology, its environmental footprint depends on how it is developed, deployed, maintained, operated and eventually, retired. Choosing modern data storage solutions such as PowerScale not only boosts AI’s efficacy in addressing sustainability challenges but also supports the sustainability of AI itself.
For more information about PowerScale, please visit our Dell AI-Ready Data Platform page and our Advancing Sustainability page for more on Dell’s efforts to promote sustainability.
1 IDC Market Presentation, Sustainable AI and AI for Sustainability – Leveraging the Upside While Mitigating the Risks of AI to Accelerate Sustainable Transformation, slide 4, Doc #US52030424, April 2024.
2 IDC Blog, Sustainable AI and AI for Sustainability, April 24, 2024.
3 IDC Market Presentation, Sustainable AI and AI for Sustainability – Leveraging the Upside While Mitigating the Risks of AI to Accelerate Sustainable Transformation, slide 7, Doc #US52030424, April 2024.
4 Based on Dell analysis comparing efficiency-related features: data reduction, storage capacity, data protection, hardware, space, lifecycle management efficiency, and ENERGY STAR certified configurations, June 2023.
