Future-proof data centers have gone from optional to foundational. As workloads expand beyond the limits of legacy facilities, enterprises need infrastructure that can support unprecedented power and cooling demands.

Lambda’s Kenneth Patchett talks with theCUBE about future-proof data centers and their role in the AI factory boom.

For legacy designs, one thing is certain: they can no longer handle the density required by modern graphics processing units, leaving an open market opportunity for hyperscalers. And this isn’t a problem where only data centers from decades back are falling behind, according to Kenneth Patchett (pictured), vice president of data center infrastructure at Lambda Inc.

“The density of hardware is such that the old data centers, meaning 2024 and behind, aren’t really suitable any longer for the current deployment of hardware,” he said. “Lambda is in the world of really moving and transforming the data center industry in such a way that we can support these large language learning models, these large training models, that are being built around the world.”

Patchett spoke with theCUBE’s Dave Vellante at theCUBE + NYSE Wired: AI Factories – Data Centers of the Future event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how GPU innovation is reshaping data center design and accelerating the push for infrastructure that can handle the demands of tomorrow.

Future-proof data centers must evolve in step with GPU innovation

The pace of hardware innovation is pushing data centers beyond their original purpose. Enterprises may want access to GPU clusters, but lack the infrastructure to sustain them, creating a gap that hyperscale providers such as Lambda are filling. This means data centers are in an architectural race to keep up with the speed of AI development, according to Patchett. 

“If you think about the rate of change, we’re getting a new piece of hardware every six to nine months, which changes the requirements to the data center that supports it,” he said. “The data centers have to keep up with the changing rate of the hardware.”

That acceleration has reshaped how providers design facilities, moving from static builds to modular systems that can develop as fast as GPUs themselves. A near-biological change, with data centers evolving in lockstep with the chips they house and working in close collaboration with partners across the supply chain, Patchett explained. 

“The way we frame it in Lambda is the data center’s DNA has to keep up with the changing DNA of the hardware,” he said. “Our ability to ensure that we’re working with our partners because … a rising tide floats all boats here. We’re working with our partners to forecast into the future what the right type of data center to build [is] to support the GPUs that are created.”

Far from theoretical, this makes for heavy work ahead for hyperscalers. In future-proof data centers, facilities must be engineered to accommodate liquid dynamics, air movement and thermal loads at levels the industry has never seen before, Patchett noted. 

“We are pushing liquid at pressures and temperatures never seen before,” he said. “There’s a huge amount of work and effort going on, and the cool thing about our industry right now is we’re all coming together and we’re underneath this problem and looking at this problem of scale, and we’re looking at it together.”

The promise behind a democratization of compute

GPU scarcity dominates industry headlines, but Patchett believes the real bottleneck lies in infrastructure. Future-proof data centers must sustain megawatt-class racks, integrating advanced cooling and power systems into a single compute engine rather than a patchwork of parts.

“Nobody wants to make a choice about researching for cancer or figuring out how to train every child in the world, like one heartbeat, one diagnosis,” Patchett said. “We are making the choices to multiply our capabilities and actually use that as an enhancement to the human existence.”

That vision ties directly to Lambda’s goal of democratizing access to compute. The company sees GPUs as democratic resources, not luxuries to hoard and envisions a world where infrastructure makes them universally available. according to Patchett. 

“We believe one GPU, one person in the world,” he said. “What happens is … the enterprise level players are able to leverage the Lambda cloud platform and they don’t have to invest their own dollars into that or create their own data center space. We’re doing that for them.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of theCUBE + NYSE Wired: AI Factories – Data Centers of the Future event:

Photo: SiliconANGLE