Click the mounts. Place your fans and GPUs. Find out whether your case really runs positive. The CFM on the box is a free-air number — a dust filter eats about a third of it, a radiator more. CaseFlow accounts for that, tracks how much your GPUs heat the case air, and shows the range instead of a single confident-looking number.
Click a fan mount on the case.
Click a GPU bay on the case.
Place some fans to get a verdict.
Saved fans live in your browser only. Nothing is uploaded.
Each fan contributes CFMmax · derate · rpm/rpmmax. The derate depends on mounting —
and on the fan: pressure-optimized blades lose ~20 % at a mesh filter, weak airflow blades up to 40 %.
| Mounting | Derate | Band | Pressure-opt. |
|---|---|---|---|
| Open grille | 0.88 | 0.80–0.95 | — |
| Dust filter | 0.70 | 0.60–0.80 | 0.78 |
| Floor + filter | 0.64 | 0.52–0.74 | 0.72 |
| Radiator 27 mm | 0.62 | 0.50–0.75 | 0.72 |
| …38 mm / 60 mm | ×0.90 / ×0.76 | thicker fins cost flow | |
| …push-pull pair | ×1.12 | series adds pressure, not parallel flow — one entry covers both fans | |
The worst-case column sums intake at the low end of each band and exhaust at the high end. If intake still wins there, the pressure holds no matter which way the estimates fall.
Case air heating: an open-cooler GPU dumps its full board power into the case air; a blower exhausts directly and leaks only ~15 %. ΔT = 3.1 · Watts / throughput (m³/h). Every degree of warmer case air is a degree of warmer air at your radiator and coolers. PSUs with their own intake chamber are a closed loop — left out.
Real-world check: a paper strip at the rear grille. Fluttering inward = negative pressure.