Featuring two of Shell’s technology partners, Penguin Solutions and AMD, Shell hosted an event at the Skybox Houston One facility in Katy, Texas, last November, where guests were able to explore cutting-edge technologies in action within Shell IT’s new high-performance computing (HPC) cluster.
Shell IT’s goal of boosting HPC capabilities while simultaneously transforming system efficiency was brought to fruition, thanks to immersion-ready systems from Penguin that feature AMD EPYC™ processors to accelerate performance paired with immersion cooling.
Attendees witnessed the 6 GRC ICEraQ Series 10 Duos, with servers fully submerged in Shell’s immersion cooling fluids, and the collective drive towards a more sustainable future for data centers first-hand. Continue reading to learn more about how Penguin, Shell and AMD are working together to boost performance and enable net-zero emissions data centers of the future.
“It’s always a pleasure, and honor, to see our solutions come to life in a production environment,” said Phil Pokorny, Chief Technology Officer (CTO) for Penguin Solutions. “The relationship between Penguin, AMD and Shell demonstrates our shared goal of reaching data center sustainability through support of technologies that can help to enable net-zero emissions operations, and we are proud to have our immersion-ready systems as an integral part of Shell IT’s HPC solution.”
“The power of collaborations like the one we celebrate with Penguin, AMD and Shell is in not only demonstrating that immersion cooling is a viable solution for data centers, but also in how we can leverage technology in a way that supports both business and sustainability goals,” said Ade Ajala, Senior Vice President of Shell Lubricants Americas.
As both an energy user and energy provider, Shell is taking on demanding compute requirements and challenges firsthand. In managing its own data centers, Shell’s IT team recognizes performance must be balanced against cost and sustainability objectives. Most recently, for its HPC cluster within its Houston data center, Shell IT identified Penguin’s Altus servers, powered by AMD EPYC processors, combined with immersion cooling technology to be an essential piece of the puzzle.
This configuration helps optimize performance relative to cost, while revolutionizing system efficiency and supporting Shell’s goal to be a net-zero emissions energy business by 2050. Continuing the drive towards sustainable data centers, Penguin Solutions, AMD, and Shell are demonstrating how a combination of renewable power and energy efficiency solutions – including immersion cooling technology – can together enable new possibilities for HPC.
The Challenge
Digital solutions are a critical component of Shell’s business, allowing the company and its customers to unlock new possibilities for cleaner energy systems, optimize existing operations, and accurately track and report emissions. But digitalization in turn, also means increasing data and workloads, requiring more energy, and impacting system performance, cost, and carbon footprint.
Shell’s Houston data center is already drawing from 100% renewable power supplied by Shell Energy North America, supporting their high-priority agenda of sustainability. The challenge for Shell’s HPC team continues to be how to drive down its Power Usage Effectiveness (PUE) ratio while simultaneously boosting performance. This motivated the upgrade to Penguin Solutions' Altus servers, powered by AMD EPYC™ processors, paired with immersion cooling technology.
The Opportunity
AMD partnered with Penguin Solutions years ago to achieve early access to new technologies, around the same time that Shell IT first leveraged AMD EPYC™ processors in their own technology. Since they had history, a natural partnership was forged, strengthening the three-way relationship and positioned the teams to deliver innovative solutions that keep the power-hungry processors cool even as processor chip wattage continues to rise.
The Technology
In its recent HPC cluster upgrade within its Houston data center, Shell IT has installed 864 dual-socket systems, using 96-core 4th Gen AMD EPYC™ 9654 GPUs, for a total of 1,728 processors and 165,888 cores. Although the overall power per rack increased, the core density of the AMD EPYC processors means it’s a much more efficient solution than air cooling which requires a data center footprint that must be spread out spatially to achieve the same performance.
CTO Pokorny concluded, “Penguin Solutions has 25 years of experience building and deploying large HPC clusters that run some of the world’s most demanding workloads. Our technology partnerships help Penguin to be at the forefront of integrating new and emerging technologies, such as immersion cooling, enabling us to meet our customers’ technology and sustainability requirements.”
To learn more, visit AMD's site, read the full partnership case study, or watch the AMD video below. Penguin Solutions offers a full array of AI & HPC infrastructure services including data center hosting and hybrid solutions as well as traditional and private cloud. Reach out today if you have outgrown your current data center environment or need to discuss design requirements. We are ready to help.
