Modularity and Units

Capacity is added as discrete burner units rather than one giant plant, so a site scales in steps that track the load it serves.

Add capacity in units

Rather than build a single enormous machine, the burner is conceived as a repeatable unit that can be deployed one or several at a time. A site's capacity is the sum of its units. This is a natural fit for the linear geometry, which repeats cleanly, and for data-center load, which grows in phases as halls are built out.

Why modular helps

• Capacity tracks demand — add units as the load grows
• Redundancy: one unit down does not take the whole site offline, aiding availability
• Learning: each unit built refines the next, following the FOAK to NOAK to BOAK path
• Maintenance can rotate through units while others keep running

Shared and independent systems

Units can share site infrastructure — cryogenic plant, grid interconnection, control room — while each keeps its own plasma, plugs, and converter. The right split between shared and per-unit systems is a design-study trade: sharing lowers duplication, independence raises resilience. The control system coordinates the units as one dispatchable block.

Modularity is also how the program manages risk honestly. The first unit is a first-of-a-kind; later units benefit from what it teaches. That staged path is why the burner roadmap runs from a test unit around 2032 toward fielded units, rather than betting everything on one build.

Standardizing on one repeatable unit also disciplines the engineering: a single design is refined and qualified rather than many bespoke builds, and the factory-like repetition that follows is what turns a first-of-a-kind machine into a fielded product line.