Think of a case fan like a garden hose. Airflow is how much water flows with the nozzle wide open. Static pressure is how hard it still pushes when you squeeze the nozzle, like air forced through a radiator or dust filter. RPM is how far the tap is turned, and noise is the hiss that comes with it. Size and thickness must fit the mount, and the connector decides how finely you can control the flow.
The width of the fan frame in millimeters, usually 120 or 140. Other sizes run from 40 to 200 mm.
When it matters: It must match the case or radiator mount. 140mm fans move more air at lower speed, so they can be quieter, but not every mount takes them.
How deep the fan frame is, in millimeters. 25 mm is standard; slim fans are around 15 mm.
When it matters: Slim fans fit tight small cases or clear tall RAM, but move less air and run louder for the same airflow. Thicker fans may not fit at all.
Skip it if: You have a standard tower with plenty of room.
How many fans come in the box.
When it matters: Compare price per fan. Packs that include a hub or controller cost more but save motherboard headers.
The distance between screw holes, in millimeters. It must match the case or radiator mounts.
When it matters: Standard 120mm and 140mm fans use set patterns. Odd sizes and hybrid frames may use different spacing, so check before buying.
Skip it if: You are buying a standard size for a standard mount.
The length of the fan's power cable, in millimeters.
When it matters: Short cables may not reach headers from top or front mounts in big cases. Extensions fix it but add clutter.
Skip it if: Your fans plug into a nearby hub.
The highest rotation speed, in revolutions per minute (RPM).
When it matters: Higher max speed gives more cooling headroom but more noise. Fans spend most of their time well below max, so compare airflow and noise at similar speeds.
The lowest speed the fan runs steadily, in RPM.
When it matters: A low minimum lets the fan stay near silent at idle. Some fans stall or stop below a certain setting.
Skip it if: You do not care about idle noise.
A reference speed for quiet use, with the airflow and noise measured at that speed.
When it matters: More useful than max figures, since most fans run slowly most of the time. Compare airflow between fans at similar noise levels.
How much air the fan moves in open space, in cubic feet per minute (CFM).
When it matters: Measured at max speed with nothing in the way. Mesh, filters and radiators cut real airflow. Makers measure differently, so compare across brands with care.
How hard the fan can push air through resistance, in millimeters of water (mmH2O).
When it matters: It matters on radiators, dense heatsinks and restrictive front panels. Pressure-focused fans suit radiators; airflow-focused fans suit open intakes and exhausts.
Skip it if: The fan goes on an open rear exhaust.
What the maker designed the fan for: case airflow, radiators, or a balance of both.
When it matters: Use airflow-tuned fans for open intakes and exhausts and pressure-tuned fans for radiators and heatsinks. Many modern fans handle both reasonably well.
How loud the fan is, in A-weighted decibels (dBA), usually at max speed.
When it matters: Maker figures use different methods and are not comparable between brands. Every 10 dBA sounds about twice as loud. Sound character, such as hum, whine or clicking, matters too.
The fan can stop fully when the controller asks it to, at low temperatures.
When it matters: It makes light use silent. Your motherboard or hub fan curve must allow a zero setting.
Skip it if: Your fans always run under load anyway.
Rubber corners or pads that stop fan vibration from reaching the case.
When it matters: They reduce hum and rattle, most noticeably on thin steel panels at higher speeds.
Skip it if: Your fans run at low speed.
What the fan spins on: sleeve, rifle, fluid dynamic (FDB), ball, or magnetic levitation designs.
When it matters: Sleeve bearings are cheapest and wear fastest, especially mounted flat. FDB and rifle suit most uses in any direction. Marketing names often rebrand these types.
The maker's estimate of fan lifespan, in hours, usually at a set temperature.
When it matters: Test conditions differ between makers, so compare within one brand. Bearing type is a better guide to long life.
Skip it if: You compare fans with the same bearing type.
What the blades are made of: standard plastic, or stiffer materials like LCP (liquid crystal polymer).
When it matters: Stiffer blades bend less at speed, which allows a tighter gap to the frame and better pressure. Mostly a premium detail.
Skip it if: You are buying mainstream fans.
Whether the fan blows the standard way or is a reverse-blade model that blows out of its front face.
When it matters: Reverse fans let intake fans show their clean side, often for bottom or side intakes in showcase cases. Check direction before buying a matched set.
Skip it if: You do not care which side of the fan shows.
The frame design: open corners, closed corners that seal against radiators, or interlocking frames.
When it matters: Closed corners seal better on radiators. Interlocking frames connect side by side and share one cable, but tie you to one brand.
Whether the fan uses 4-pin PWM (pulse width modulation) for precise speed control, rather than 3-pin voltage control.
When it matters: PWM gives smoother, lower speed control. Either plug fits either header, but a 3-pin fan on a PWM-only header may run at full speed.
The fan's power plug: 3-pin, 4-pin PWM, or a brand's own plug for its hub.
When it matters: Brand plugs need that brand's controller. Some cheap fans use a Molex plug and always run at full speed.
The most power one fan uses at full speed, in watts.
When it matters: Add up fans sharing one header or splitter. Typical headers are rated around 1 amp (12W); going over risks damage. A powered hub avoids the limit.
Skip it if: Each fan has its own header.
The lowest voltage at which the fan starts spinning from rest.
When it matters: Matters for 3-pin fans under voltage control. If the header's minimum sits below this, the fan may not start at low settings.
Skip it if: You use PWM control.
The lowest PWM setting, in percent, that still changes the fan's speed.
When it matters: Helps when setting fan curves: settings below the floor change nothing, or stop the fan on zero RPM models.
Skip it if: You leave fan curves on default.
Fans link to each other so a group runs from one cable.
When it matters: It cuts cable clutter and saves headers. Chains have a fan limit, and header current limits still apply. It usually works only within one brand's series.
Skip it if: You are installing one or two fans.
Whether the fans need the brand's hub or software to work, and which product family they belong to.
When it matters: Closed ecosystems tie future fans, pumps and lighting to one brand and app. Mixing brands can mean extra hubs and apps.
Skip it if: The fans use standard 4-pin plugs.
The plug for fan lighting: 3-pin 5V ARGB (addressable), 4-pin 12V RGB, or a brand's own plug.
When it matters: 5V ARGB and 12V RGB are not interchangeable. Plugging ARGB into a 12V header can kill the LEDs. Check your board or hub headers first.
Skip it if: The fan has no lighting.
Whether and how the fan lights up: none, single color, RGB, or addressable RGB with effects per LED.
When it matters: Addressable lighting needs a matching header or hub plus software. Lighting adds cost, not cooling.
Skip it if: You want plain fans.