Specifications
Family
SmartPodX · SmartPodXL · SmartPodXL+
Cooling Type
Single-phase immersion (open tank)
Heat Dissipation — SmartPodX
50 kW
Heat Dissipation — SmartPodXL
50 kW (single CDU)
Heat Dissipation — SmartPodXL+
100 kW (dual CDU)
IT Capacity — SmartPodX
21U / 19 OU
IT Capacity — SmartPodXL
44U / 42 OU
IT Capacity — SmartPodXL+
41U / 39 OU
Dimensions — SmartPodX
120 (L) × 90 (W) × 119 (H) cm
Dimensions — SmartPodXL / XL+
231 (L) × 93 (W) × 121 (H) cm
Weight Empty — SmartPodX
411 kg
Weight Full — SmartPodX
872 kg (with SmartCoolant)
Weight Empty — XL / XL+
671 kg / 767 kg
Weight Full — XL / XL+
1,620.5 kg / 1,716.5 kg
Coolant
Submer SmartCoolant synthetic dielectric
Coolant Fire Point
233 °C
Coolant Pour Point
−50 °C
PUE (with secondary cooling)
≈1.03
Cooling Power vs Air
≈30 W per kW vs ≈670 W per kW
Overview
SmartPodX, SmartPodXL and SmartPodXL+ are three sizes of the same open-tank single-phase immersion platform, covering 50 kW to 100 kW of heat dissipation. The X takes 21U of 19-inch or 19 OU of OCP ORv3 gear in a short 120 cm body; the XL and XL+ stretch to 231 cm to take 44U / 42 OU and 41U / 39 OU respectively, with the XL+ adding a second CDU to reach 100 kW.
Because the platform is shared, a site can start with a SmartPodX and scale to XL+ units without changing the fluid, the monitoring tooling or the operating procedure — the tanks differ in volume and CDU count, not in principle. Filled weights matter for floor loading: 872 kg for a filled X, around 1.62 t for a filled XL and 1.72 t for an XL+.
Submer's published comparison against air cooling is the commercial argument: roughly 30 W per kW of cooling power versus about 670 W per kW, and a PUE near 1.03 once secondary cooling is included. Immersed hardware is also reported to live 20–50 % longer, because there are no fans, no airborne contaminants and far smaller thermal excursions.
Key Benefits
Scale on one platform — 50 kW to 100 kW across three tank sizes with the same coolant and tooling. OCP ORv3 native — 19 OU to 42 OU of Open Compute gear without adapters. Measurable efficiency — ≈30 W per kW of cooling power against ≈670 W per kW for air.
Applications
Colocation and enterprise halls adopting immersion, HPC and AI clusters between 50 kW and 100 kW per tank, retrofit of existing racks, and hydrogen or heat-reuse projects that need a predictable warm-water return temperature.
Request a Quote — SUBMER SMARTPODX / XL / XL+ — 50–100 KW IMMERSION POD FAMILY
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →