Tropical · the honest version

Liquid cooling in the tropics — the real problem is where the heat goes.

25 °C30 °C35 °C — CAPACITY ≈ HALFWET-BULB 30 °C = NO FREE COOLING

Most of the world’s next data center capacity is being built in warm climates, where the physics is least forgiving. We build for those rooms — and we will not pretend the tropics are free.

Physics, not marketing

Three facts about heat in warm rooms

≈ ½

Capacity halves at 35 °C

An L2A loop rated at 25 °C room air delivers roughly half at 35 °C. Every L2A product on earth obeys this; not every vendor publishes it. We do.

1.1×

The room still has to drink the heat

Tropical deployment does not remove the building AC — it borrows it. The room must extract ≥ 1.1 × IT load. The worksheet is public.

30 °C

Outdoor rejection is mostly dead

At tropical wet-bulbs, dry coolers and free cooling collapse — mechanical refrigeration carries the load. "Independent of building AC" claims should be read accordingly.

Standards context

SS 726:2026 — the year the tropics got a standard

What it is

Singapore’s SS 726:2026 (Aug 2026) — the first national standard for liquid cooling in tropical climates. A milestone for the whole category: the tropics stopped being an edge case.

What it is not (for us, yet)

Its scope is data-center-internal, direct-to-chip, single-phase systems — our standalone edge cabinet sits outside that scope. We reference SS 726 as context, and we do not claim compliance. When a scope fits, we will test against it and publish the result.

The killer asset, in progress

35 °C full-load derating curve — measured, not modeled

No vendor in this category publishes a measured capacity-vs-room-temperature curve at full load. We are building the demo unit precisely so this page can carry one: 4-hour sustained 100% load at 35 °C, raw data, including any throttle events. Until then, this page shows the planning envelope derived from published L2A physics — and says so.

Size it for a warm roomFull hot-climate derivation ↗