Earth Day 2024: HPE partnering with customers to achieve sustainability in the AI age
Despite being an energy-intensive workload, AI can play a pivotal role in integrating sustainability into our mission, driving innovation, and fostering growth
- The HPE Sustainability Insight Center, a component of the HPE GreenLake platform delivers actionable insights around energy consumption, carbon emissions, and electricity costs
- Long-time customer, Danfoss, shares how the company’s AI solutions are integral to its decarbonization innovations and the importance of HPE GreenLake offerings for supporting the company’s sustainability mission
Roles played in accomplishing business and energy goals
As Earth Day 2024 approaches, I have been considering how HPE and our customers contribute to climate solutions while simultaneously achieving our business objectives. Many of us are taking action to make sustainability a core part of our mission to spur innovation and growth.
This is particularly important as we fully enter the AI age, which holds incredible promise for transforming our world while also introducing significant decarbonization challenges.
Take, for example, GPT3, which was the precursor of ChatGPT. Training the Large Language Model (LLM), consumed the energy equivalent of 30 average U.S. homes for an entire year. More recent LLMs have doubled or tripled that power consumption. Further, the energy required for training an LLM is only about a tenth of the total power budget. The inferencing phase for an AI model consumes at least 10 times more energy.
I believe it’s numbers like these that create a mental barrier to adopting more aggressive sustainability goals: There is a perceived conflict between business growth and decarbonization, as it seems to make sense that growth requires a corresponding increase in power usage and carbon emissions.
Designing and deploying intelligent, high-availability, low-power computing infrastructure is critical to containing the cost and energy requirements for utilizing AI to solve business problems and create better futures for all
Harnessing the power of AI to control carbon emissions and energy consumption
But this need not be the case, as technology can also offer solutions. It’s why designing and deploying intelligent, high-availability, low-power computing infrastructure is critical to containing the cost and energy requirements for utilizing AI to solve business problems and create better futures for all.
It’s also why we take sustainability incredibly seriously as we continue evolving the HPE GreenLake platform and portfolio of as-a-Service offerings by engineering end-to-end sustainability innovations, like liquid cooling, into the platform.
These innovations include fully rolling out the HPE Sustainability Insight Center as a component of the HPE GreenLake platform. Plus, we’re empowering organizations with AI-enabled tools that help them discover and monitor their entire hybrid cloud estate by including OpsRamp as part of every new HPE GreenLake deployment. This ensures organizations can continuously optimize their IT environments for energy efficiency.
Using the Sustainability Insight Center and OpsRamp, organizations gain key insights on three important IT sustainability factors—energy consumption, carbon emissions, and electricity costs—across a company’s entire IT estate, regardless of what vendor is supplying the hardware and software.
In other words, the Insight Center analyzes consumption, emissions, and power costs across multi-vendor and multi-cloud estates. Then it delivers insights via an intuitive dashboard that enables IT staff to determine how workloads can be redistributed and optimized for improved sustainability
In other words, the Insight Center analyzes consumption, emissions, and power costs across multi-vendor and multi-cloud estates. Then it delivers insights via an intuitive dashboard that enables IT staff to determine how workloads can be redistributed and optimized for improved sustainability.
Danfoss: Leading the way on climate and AI
One of our customers leading the way in sustainability and AI is the global engineering and manufacturing enterprise Danfoss. With decarbonization at the core of its mission, Danfoss creates innovative solutions that span energy generation and delivery, cold chains, automative applications, industrial equipment, integrated building controls, and others.
At the recent HPE Discover in Barcelona, Danfoss CIO Sune Baastrup joined me on stage to discuss the ways his company is applying AI and other technologies to decarbonize some of the heaviest industries in the world. There, we learned how Danfoss leverages multiple instances of HPE GreenLake to help it with embedding AI and machine learning deeply into its value chain.
For example, Danfoss has engineered a solution for district energy suppliers that makes heating and cooling homes, business, and factories more efficient. It utilizes AI to analyze temperature and other data collected by hundreds of thousands of sensors, coupled with external sources of weather data, to determine the optimum ways for pre-heating and pre-cooling structures. The solution significantly reduces energy utilization while delivering superior comfort.
Decarbonizing IT with HPE GreenLake
An early adopter of HPE GreenLake for Private Cloud Enterprise, Danfoss is applying this same decarbonization mindset to its IT organization. The partnership has been critical to enabling Danfoss to continue its sustainability leadership while undergoing exceptionally strong business growth.
In collaboration with HPE and others, Danfoss is optimizing the energy ecosystem for its data centers that has cut up to 50% of the cooling energy required, compared with traditional cooling technologies. It also reuses the excess heat coming out of those data centers for various purposes, including covering significant percentages of the power required at its nearby manufacturing facilities.
As a data-driven company, the addition of the HPE Sustainability Insight Center is providing Danfoss with new visibility into energy utilization across all of its IT systems. This is enabling IT staff to drive further energy efficiencies within its extensive technology estate and work with all of its partners to reduce power consumption and carbon emissions.
Danfoss also relies on HPE for asset upcycling as part of its HPE GreenLake deployment. Implemented throughout multiple geographies, Danfoss established a secure, systematic, and sustainable process for retiring IT assets through HPE Asset Upcycling Services. This enables the company to recover economic value from end-of-life equipment while ensuring responsible disposal that includes best‑in-class chain of custody to help assure the safe handling and consistent asset disposition across its global operations.
Trusted, long-term relationships key to sustainable AI
Moving ahead, Danfoss expects to expand its AI development partnership with HPE, such as extending its HPE GreenLake utilization to further enhance the solution’s decarbonization benefits. Options include HPE GreenLake for Large Language Models, and other recent data center innovations, for placing high-performance, high-availability computing resources in close proximity to data and AI systems for maximizing speed while minimizing power consumption.
As Sune shared at Discover, strong, trusted, long-term relationships are key to successfully forging new solutions that incorporate AI technologies while staying true to our sustainability and decarbonization objectives.
We look forward to our continuing relationship with Danfoss, and our many other customers that have helped us hone our solutions to make them more sustainable. We appreciate and value their support of our mutual vision for achieving carbon neutrality across our enterprises.
To find out how you can advance your decarbonization objectives and goals, we invite you to learn more about HPE GreenLake, HPE GreenLake for Large Language Models the HPE Sustainability Insights Center dashboard.