Blog
NGC2 Doesn’t Operate in a Data Center
Why tactical edge infrastructure must be engineered for the extreme environments of NGC2
The Army’s NGC2 effort is fundamentally about information: moving data across formations, connecting sensors and applications, enabling AI, and giving commanders the information they need to make decisions faster.
But every digital capability in that architecture ultimately depends on something physical.
Servers have to process the data. Network equipment has to move it. Storage has to retain it. Power has to remain available. And all of that equipment has to keep operating wherever the mission takes it.
That distinction matters because NGC2 is not being built for a climate-controlled data center. It is being pushed toward the tactical edge-into vehicles, command posts and distributed formations operating through heat, dust, vibration, unreliable power and degraded connectivity.
Recent Army experimentation has made that physical reality increasingly visible.
At Project Convergence Capstone 6 (PC-C6) at Fort Irwin, California, roughly 10,000 participants experimented with more than 90 technologies, concepts and formations under harsh environmental conditions. The Army described the event as its largest Project Convergence iteration to date and an opportunity to evaluate NGC2 at division scale.
The lesson extends well beyond any individual piece of equipment: Moving computing closer to the mission changes what computing has to survive.
From experimentation to the field
NGC2 represents a significant change in how the Army approaches command and control.
Instead of relying on a collection of fragmented systems built around individual functions, the Army describes NGC2 as an architecture intended to unify data across warfighting functions and allow applications and AI capabilities to operate across echelons.
In June, the Army reported that it had advanced NGC2 from proof of concept to division-scale experimentation in less than a year. By July, PC-C6 was putting that architecture through a much more demanding test at the National Training Center.
Reporting from PC-C6 described hardware overheating in the Mojave Desert and soldiers encountering intermittent connectivity. Army leaders acknowledged that some technologies had difficulty handling the heat while emphasizing that finding those friction points was part of the purpose of operational experimentation.
A system can perform exactly as expected on a test bench and still encounter entirely different challenges once it becomes part of a moving tactical formation. The field changes the engineering problem.
The environment changes the requirements
Consider what conventional computing infrastructure normally assumes.
A data center provides controlled temperatures, stable utility power, predictable airflow, fixed installation and persistent high-bandwidth networking. Equipment operates inside an environment deliberately engineered around the needs of the hardware.
At the tactical edge, that relationship is reversed. The hardware has to be engineered around the environment.
Temperatures can rise dramatically during the day and fall at night. Fine dust and sand can challenge cooling systems and connectors. Equipment mounted in vehicles is subjected to continuous vibration and shock. Power may come from vehicles, batteries, generators or other DC sources and may not always remain stable.
The network itself becomes another variable. Units may operate through denied, degraded, intermittent and low-bandwidth – or DDIL -conditions rather than the persistent connectivity expected in an enterprise environment. The Army’s current NGC2 fielding describes infrastructure intended to support real-time data exchange under exactly those conditions.
A server that works perfectly when installed permanently in a rack solves a different problem than a compute node that must travel with a formation, operate from a vehicle and repeatedly deploy, move and redeploy.
These are not secondary considerations surrounding compute performance. They are part of compute performance.
Compute is becoming physical infrastructure for NGC2
It is easy to think about NGC2 primarily in terms of applications, AI and data. But those capabilities sit on top of a physical stack.
At a simplified level, the architecture can be viewed as four interconnected layers: Applications → Data & AI → Infrastructure → Transport
Applications depend on data. Data and AI workloads depend on available computing and storage. Computing depends on transport, networking and power. Remove the physical foundation and everything above it becomes unavailable.
That is why the Army’s transition from experimentation toward wider fielding makes infrastructure increasingly important. In September, III Armored Corps began receiving NGC2 baseline capabilities starting with the transport and infrastructure layers. Army officials described that network transport and compute infrastructure as the baseline for current and future modernization efforts.
What Changes at the Tactical Edge
Moving compute closer to the mission enables local processing, AI and applications—but introduces new physical demands.
Thermal Management: Compute must sustain performance through extreme temperatures.
Shock & Vibration: Hardware must withstand vehicles, transport and rough terrain.
Resilient Networking: Systems must remain useful through intermittent and DDIL connectivity.
Power Continuity: Edge infrastructure must accommodate variable DC power and outages.
Mobility: More capability has to fit into smaller, deployable footprints.
Local Processing: Critical workloads must remain available without constant reach-back.
These factors determine whether performance remains available when conditions deteriorate.
Extreme Conditions Aren’t Edge Cases
The Mojave heat encountered during PC-C6 highlights only one challenge. Tactical infrastructure may also face humidity, salt exposure, extreme cold, dust, vibration and unstable power.
The environment changes, but the requirement remains the same: tactical infrastructure must be engineered for where it will actually operate.
Ruggedization Starts with the System
Ruggedization is more than placing commercial hardware inside a protective enclosure. Thermal design, chassis construction, components, connectors, power and networking all influence whether compute remains available in the field.
The goal is not simply surviving the environment. It is maintaining useful computing capability throughout the mission.
Designing the Physical Edge
This is the challenge platforms such as the Core Systems ATMOS2 Series are designed around. ATMOS2 edge nodes combine rugged compute, networking and integrated power in a modular, stackable architecture built to move processing closer to tactical users.
As NGC2 moves down echelon, infrastructure must become smaller, more resilient and less dependent on data-center conditions.
In Conclusion
NGC2 may be driven by data, applications and AI, but those capabilities depend on physical infrastructure. Real-world operations introduce heat, dust, vibration, disrupted connectivity and variable power, all factors that can determine whether compute is actually available when needed.
At the tactical edge, performance is only valuable if it remains available where the mission happens.
