Applied Digital’s latest move is turning heads in the HPC sector. The company just broke ground on a landmark high-performance computing facility in Ellendale, North Dakota—and the specs are nothing short of impressive.
The Technical Game-Changer
Here’s what makes this 342,000-square-foot facility stand out: it’s engineered from the ground up to handle today’s most demanding AI workloads. We’re talking generative AI, natural language processing, machine learning, and heavy-duty rendering. The design philosophy? Maximize compute density while keeping cooling and power distribution efficient.
The numbers tell the story. Applied Digital is fitting roughly 50,000 Nvidia H100 SXM class GPUs into a single compute cluster using proprietary high-density rack configurations. That’s a density game-changer in an industry where space and power are everything.
“We literally had to think inside a smaller box,” explains Brad Barton, EVP of Real Estate Development at Applied Digital. The team designed the internal network infrastructure first, then built the power and cooling systems around it—a deliberate reversal of conventional datacenter architecture.
Why Location Matters
North Dakota’s abundant renewable power sources give Applied Digital a significant edge on operational costs. Combined with the company’s innovative liquid-cooled infrastructure design, customers get a rare combination: lower costs paired with superior performance. That’s particularly compelling for enterprises racing to scale their AI capabilities without breaking the bank.
The Market Opportunity
The timing aligns perfectly with explosive demand. The global HPC market is projected to hit $65.12 billion by 2030, with AI-related workloads driving most of that growth. Applied Digital’s purpose-built approach positions them to capture a meaningful slice of this expansion.
“The AI landscape is evolving rapidly, and compute power needs are scaling accordingly,” notes Mike Maniscalco, CTO of Applied Digital. “Our unique design and strategic infrastructure placement create something we haven’t seen before at this scale.”
As the industry races to build AI-capable infrastructure, facilities like this Ellendale installation represent the next generation of datacenter thinking—one where every architectural decision is optimized for the workloads of the future.
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
Applied Digital Raises the Bar: New 100MW AI Infrastructure Hub Takes Shape in North Dakota
Applied Digital’s latest move is turning heads in the HPC sector. The company just broke ground on a landmark high-performance computing facility in Ellendale, North Dakota—and the specs are nothing short of impressive.
The Technical Game-Changer
Here’s what makes this 342,000-square-foot facility stand out: it’s engineered from the ground up to handle today’s most demanding AI workloads. We’re talking generative AI, natural language processing, machine learning, and heavy-duty rendering. The design philosophy? Maximize compute density while keeping cooling and power distribution efficient.
The numbers tell the story. Applied Digital is fitting roughly 50,000 Nvidia H100 SXM class GPUs into a single compute cluster using proprietary high-density rack configurations. That’s a density game-changer in an industry where space and power are everything.
“We literally had to think inside a smaller box,” explains Brad Barton, EVP of Real Estate Development at Applied Digital. The team designed the internal network infrastructure first, then built the power and cooling systems around it—a deliberate reversal of conventional datacenter architecture.
Why Location Matters
North Dakota’s abundant renewable power sources give Applied Digital a significant edge on operational costs. Combined with the company’s innovative liquid-cooled infrastructure design, customers get a rare combination: lower costs paired with superior performance. That’s particularly compelling for enterprises racing to scale their AI capabilities without breaking the bank.
The Market Opportunity
The timing aligns perfectly with explosive demand. The global HPC market is projected to hit $65.12 billion by 2030, with AI-related workloads driving most of that growth. Applied Digital’s purpose-built approach positions them to capture a meaningful slice of this expansion.
“The AI landscape is evolving rapidly, and compute power needs are scaling accordingly,” notes Mike Maniscalco, CTO of Applied Digital. “Our unique design and strategic infrastructure placement create something we haven’t seen before at this scale.”
As the industry races to build AI-capable infrastructure, facilities like this Ellendale installation represent the next generation of datacenter thinking—one where every architectural decision is optimized for the workloads of the future.