Liquid Cooling, the Answer to China's Token Explosion

From 10 to 3,900 Quadrillion Tokens by 2030: A 370-Fold Inference Surge and the Liquid Cooling Curve That Follows It

JPMorgan forecasts China's AI inference token consumption rising from about 10 quadrillion in 2025 to roughly 3,900 quadrillion by 2030, a near 370-fold jump, with ByteDance's Doubao passing 50 trillion daily tokens, and token growth pushes data centers toward high density where air cooling cannot follow, leaving liquid cooling as the key support of AI compute infrastructure. Ten domestic companies ride the curve: Shangluo leads the gainers at 323.74%, Runze batch-delivers across 320,000 planned racks, Envicom lifts liquid cooling share from 8% to 35%, and Shenling ranks first among CDU vendors with 72 patents. EXTRCOOL provides one-stop liquid cooling solutions.

More information

Details

Area 51, NV, USA

EXTRCOOL INDUSTRY LIMITED

Associated Trend items

Surface Treatment for Liquid Cooling Manifolds, the Natural-Finish Pickling Process

Matte White to Natural Luster, Blue-Point From 12-18 to Above 20 Seconds: The Two-Stage Route That Removes Weld Scale Without Touching the Base

Surface treatment for liquid cooling manifolds starts from the weld problem: the heat-affected zone triggers intergranular chromium depletion that degrades surface oxidation and roughness, threatening long-term corrosion resistance, and conventional pickling removes the scale but turns the weld grey, rough and the stainless matte white. The natural-finish route runs two ...

More information

The Stainless Manifold, the Hidden Gatekeeper of AI Server Liquid Cooling

One Uneven Split, One Throttling GPU: Why the 120-140kW NVL72 Turned the Manifold Into NVIDIA's Certified Gatekeeper

The stainless manifold is the hidden gatekeeper of AI server liquid cooling: the cold plate sits on the GPU, the CDU handles exchange, the manifold splits one flow into dozens and the quick coupling allows maintenance without breaking the loop, so in an NVL72 cabinet an uneven split lets one GPU drift hot, throttle first or trip protection. Stainless wins on ten-plus-yea...

More information

Manifold and Quick Coupling Selection, the Field Guide for Cold Plate Cooling

0.86 × Heat ÷ ΔT With 20% Margin, Server Thread First, 20 Bar and 5,000 Cycles: The Five-Step Selection That Stops Leaks, Not Just Flow

Manifold and quick coupling selection for cold plate cooling starts with the failure mode: the manifold divides one flow into many branches but its risk lives in the interfaces, so seal reliability outranks flow testing. Sizing follows 0.86 times the heat over the temperature difference with 20 percent margin, a 40 by 40 tube holding 75 liters per minute for 25 to 50 kil...

More information

Products associated

modular micro datacenter

EXTRCOOL 20FT

modular data center

LL-Block40

Compare

Remove all

Compare up to 10 products