- If your AI-generated code becomes faulty, who faces the most liability exposure?
- These discoutned earbuds deliver audio so high quality, you'll forget they're mid-range
- This Galaxy Watch is one of my top smartwatches for 2024 and it's received a huge discount
- One of my favorite Android smartwatches isn't from Google or OnePlus (and it's on sale)
- The Urgent Need for Data Minimization Standards
Improving Data Center Energy Efficiency to Power the AI Revolution
The world’s energy grids were not designed to support hyperscale data centers and pervasive cloud services, let alone artificial intelligence (AI) workloads. With AI workloads becoming more commonplace and scaling at unprecedented rates, energy grids are more strained than ever before.
“The public cloud created a wave of change in how data centers are built and operated,” said Vladimir Ester, CTO and Cofounder of ClusterPower, during a February 2024 panel discussion on IT sustainability. “And in the past couple of years, we’re seeing the next revolution—which is being generated by AI deployments and, more recently, very large-scale Gen AI-related infrastructures.”
This AI revolution, he added, is contributing to the global energy crisis.
AI, which requires more servers, storage, processing power, and cooling than other types of workloads, has historically been considered at odds with sustainability. But progress is now being made on multiple levels.
“There’s a need to become more sustainable and efficient at scales that weren’t possible before,” according to Ester.
Rules and regulations are being established that are driving data centers to be more energy efficient and advance sustainability, and the computing platforms that fuel those data centers are playing an active role in enabling these improvements.
Driving efficiency and sustainability at stakeholder, facility, and platform levels
Aiming to be climate-neutral by 2050, the European Union (EU) is confronting the energy crisis directly. EU standards, requirements, and other initiatives—such as the EU Code of Conduct for Data Centres, the EU Energy Efficiency Directive (EED), and EN 50600—are aimed at improving energy efficiency, reducing energy consumption, and advancing sustainability.
Not only are these regulations compelling enterprises to rethink and revise their operations, but they are also spurring innovation and new business models.
ClusterPower, for example, built the largest data center facility in Romania with sustainability in mind. It is the first in Southeastern Europe with dual Tier III certifications, and has its own microgrid that produces energy from gas and hydrogen. The facility even recovers heat generated from its technology systems, which is then used to cool those same systems via an innovative process. The data center’s foundational computing infrastructure is built on Cisco UCS X-Series Modular System servers which, according to internal testing, can support:
- Up to 70% reduction in total footprint compared to previous generation servers
- Up to 49% reduction in total power consumption with M7 systems compared to M4 systems
- Up to 90% reduction in hardware operating costs
- Up to 72% reduction in hardware maintenance costs
- Up to 75% reduction in recurring software support costs*
In addition to winning multiple awards related to energy efficiency and sustainability, the modularity of the UCS X-Series supports the performance, flexibility, and scalability Ester said is needed for today’s workloads—as well as tomorrow’s larger, more power-intensive workloads. And the multitenancy and microsegmentation capabilities of ClusterPower’s Cisco Application Centric Infrastructure (ACI)-based network are driving operational consistency and efficiency while maintaining isolation and protection of multiple client environments.
“Efficiency is extremely important. We strive to do more with less and we need to optimize the cost of our operations,” Ester said during the panel discussion. “The AI revolution and its challenges are also extremely important. It’s one thing to deploy a general cloud or computing platform; it’s a completely different thing to add a bunch of GPUs that need to be scaled quickly. My advice for those who are building new data centers would be to come to leaders like Cisco, Intel, and ClusterPower, and look at models that have already been deployed and validated by the industry before jumping in and making investments.”
To learn more about ClusterPower’s data center and use of Cisco
UCS X-Series and Cisco ACI, read the full case study.
*Cisco Networking Compute Executive Briefing Center: X-Series Sustainability, January 2024
Share: