Thermal-buffer modules and high-conductivity coolant additives engineered for AI workload thermal profiles — peak shaving, demand-charge offset, and recovery-window extension for high-density compute.
1. Density. Rack thermal loads at 50–120 kW are no longer outliers. Air-side cooling alone struggles to remove that heat at acceptable approach temperatures.
2. Volatility. Training jobs and inference traffic produce sharp thermal peaks that drive peak demand charges and short-cycle compressors.
3. Resilience. Cooling outages and chiller transitions create narrow recovery windows. Short interruptions can translate into expensive workload pauses.
Most facilities respond by oversizing — more chiller capacity, more redundancy. Thermal-buffer capacity is the alternative axis. A buffer absorbs the peaks, smooths the duty cycle, and extends recovery windows without committing to permanent additional cooling capacity.
Container-scale thermal buffer, sited next to the load — absorbing peaks without permanent added chiller capacity.
Two integration paths — depending on whether you want to specify your own thermal subsystem around our materials, or take a pre-engineered module from us.
Specify Passive Edge materials into your own thermal subsystem. Use UltraST PCM as the storage medium and nano-coolant additives to lift conductivity of working fluid.
Pre-engineered modules built around UltraST cores, with integration interfaces matched to typical liquid-cooling and rear-door heat-exchanger topologies. Validated per project.
The more of the following you can share early, the faster we can return a useful response. None of these constitute a commitment from either side.
We work in clearly-bounded stages. No stage commits the next. Each stage produces a document you can use internally — even if you don't proceed with us.
You share load profile, topology, tariff, and target outcome. We return a written fit assessment within ~5 business days.
Joint review of which Passive Edge material or module is the best fit, with technical data package and integration spec attached.
Scoped pilot — typically rack- or row-level — with success criteria, instrumentation, and a defined validation window.
Pilot data review and engineering recommendation. Decisions about scaled deployment are made on validated evidence.
Phase point selection, latent heat capacity, conductivity, cycle stability data, encapsulation chemistry.
Integration patterns for rear-door, direct-liquid, and immersion topologies. Sizing worksheets and pilot scaffolds.
Editable template covering instrumentation, success criteria, validation windows, and reporting outputs.
Our team will respond with a fit assessment, a relevant technical data package, and a proposed pilot test plan if there is a match.