Consulting & Audit

We begin every engagement with an honest assessment of whether immersion cooling is the right fit for your facility, workload, and constraints.

Site Assessment

On-site review of floor loading capacity, existing HVAC and electrical infrastructure, heat rejection options, and facility layout. We identify constraints early that influence deployment viability.

Thermal & Workload Analysis

Analysis of current and projected heat densities (kW per rack), workload profiles, and growth trajectory. We model thermal requirements to size systems correctly.

TCO Modelling

Total cost of ownership comparison: immersion vs continued air cooling. Typical payback ranges: 2–5 years depending on scale, energy costs, and facility constraints.

System Design

Custom architecture tailored to your facility, workload density, and operational requirements.

Tank & Rack Architecture

Selection and sizing of immersion tanks — typically 4–20 nodes per tank. Rack layouts, cable management, and maintenance access designed for operational efficiency.

Fluid Selection

Vendor-neutral evaluation of dielectric fluids: mineral oil, synthetic esters, and engineered two-phase fluids. Consideration of thermal performance, longevity, and Australian supply chain.

Heat Rejection Design

Secondary cooling loop design: dry coolers, adiabatic systems, or chiller integration. Optimised for Australian climate zones. Typical approach temperatures: 3–8°C above ambient.

Deployment & Build

Procurement, fabrication, installation, and commissioning. Turnkey or assisted-build options.

Procurement

Vendor-neutral sourcing of tanks, fluid, heat exchangers, pumps. Typical lead times: 4–12 weeks depending on equipment type and supply chain conditions.

Installation

Assembly and placement of tanks, plumbing, heat rejection, electrical tie-ins. Typical pilot deployment: 6–8 weeks from equipment arrival to first workload.

Commissioning

Fluid fill, leak testing, pump and control verification, thermal validation. Baseline PUE measurement, operational documentation, and staff training included.

Operations & Maintenance

Ongoing support for mission-critical immersion cooling systems.

Monitoring

Integration with DCIM and BMS. Fluid temperatures, flow rates, pump status. Optional 24/7 NOC support for critical deployments.

Fluid Management

Periodic fluid quality testing, top-up procedures, replacement schedules. Typical single-phase fluid life: 5–10+ years with proper maintenance.

Preventive Maintenance

Scheduled inspections of pumps, seals, filters, heat exchangers. Quarterly for standard deployments; monthly for high-availability environments.

Where Immersion Cooling Fits

Representative deployment scenarios where immersion delivers clear advantages over traditional cooling.

AI Training Clusters

GPU clusters (H100, B200, and beyond) generate 50–100+ kW per rack. Air cooling cannot support these densities; liquid cooling becomes mandatory. Immersion offers no per-node plumbing, simplified maintenance, and tolerance for dense GPU-to-GPU topologies. Typical deployments: 8–64 GPUs per tank.

HPC & Research

Universities, CSIRO, and research institutions running simulation, weather modelling, genomics, or CFD at 20–40 kW per rack. Immersion reduces cooling energy by 40–50% and increases density for expansion within constrained space. Stable thermal environments support reproducible results.

Defence & Government

Data sovereignty, secure supply chains, and deployability in austere environments. Self-contained tanks reduce reliance on facility HVAC; lower acoustic signatures benefit tactical deployments; Australian-owned consulting avoids foreign dependencies for sensitive infrastructure.

Enterprise Data Centres

Reduce PUE, extend hardware life, and increase density without building new facilities. Immersion retrofits deploy in existing space — floor loading is the primary constraint. Target: 30–50% cooling energy reduction and 50+ kW per rack support.

Edge & Remote Deployments

Mining sites, offshore platforms, remote telecommunications, and regional facilities where traditional HVAC is impractical. Self-contained immersion units: no CRAC, reduced electrical load, operation in harsh environments. Container-based solutions available.

Immersion vs Air Cooling

Typical ranges for comparison. Actual performance depends on facility design, climate, and workload.

MetricImmersion CoolingAir Cooling
PUE (typical)1.02 – 1.101.3 – 1.6
Rack density50 – 100+ kW per rack10 – 20 kW per rack
Cooling noise~60 – 65 dB (pumps only)~75 – 85 dB (fans, CRAC/CRAH)
Floor footprintHigher density = less m² per kWLarger footprint for same compute
MaintenanceFluid testing, pump/filter serviceFilter changes, CRAC service, ductwork
Facility requirementsFloor loading, fluid plumbing, heat rejectionRaised floor, HVAC, ductwork
Hardware compatibilityImmersion-ready servers (OEM or retrofit)Standard servers

Values are indicative. Site-specific assessment recommended.

Ready to Explore Immersion Cooling?

Tell us about your infrastructure challenges. We'll provide an honest assessment of whether immersion cooling is right for your deployment.