Passive vs active GPU cooling

A passive GPU cooler is a fanless heatsink that relies on the server's chassis fans to push air through it. An active GPU cooler carries its own fan on the card. NVIDIA's nvidia-smi tool reports no fan speed for many parts, since they rely on fans in the surrounding enclosure. The RTX PRO 6000 Blackwell Server Edition is one such passive card.

NVIDIA builds the RTX PRO 6000 Blackwell in three dual-slot editions. The Server Edition is passive at 400 to 600 W. NVIDIA labels the 600 W Workstation Edition's cooler double flow-through and the 300 W Max-Q's cooler active. For the Server Edition, NVIDIA lists a minimum of 65 cubic feet per minute (CFM) of ducted airflow at a 35 °C heatsink inlet. At 45 °C it lists 125 CFM, and 43 CFM applies whenever the card has power, even at idle. Short of that airflow, the card can reach its hardware slowdown temperature, which halves core clocks.

Install a passive card only in a chassis whose maker qualifies it, since NVIDIA's airflow figures assume fully ducted air. For a desktop tower, choose an active card such as the RTX PRO 6000 Max-Q, which NVIDIA rates for up to four cards per workstation.

Sources

Last verified October 7, 2026.