Thin gaming laptops trade thermal headroom for portability, so your fan curve matters more than it does on a thick chassis. With less heatsink mass and tighter fan blades, heat builds fast and noise ramps aggressively. The goal is not to keep the machine silent at all costs or to blast fans at 100% constantly, but to match airflow to actual load with enough delay and hysteresis to avoid constant revving. Prioritize smooth ramping, separate CPU and GPU control where possible, and a temperature-to-speed relationship that keeps idle quiet, handles sustained gaming without thermal throttling, and avoids unnecessary wear. The stock curves on most thin models are either too passive until 90C or too loud at 50C. A custom curve fixes both.
As of early 2026, most control apps have standardized on a graphical point-based editor, which makes it practical to dial in precise behavior without BIOS editing for most users.
As an Amazon Associate we earn from qualifying purchases at no extra cost to you.
Top 3 picks at a glance
Quick Recommended Settings
| Setting | Recommended value | Why |
|---|---|---|
| Control Software | FanControl v195+ or Manufacturer app (G-Helper / MSI Center / Armoury Crate) | Gives per-sensor control and hysteresis that Windows power plans do not |
| CPU Fan Curve (Thin Chassis) | 30C=0%, 45C=25%, 60C=45%, 75C=70%, 85C=90%, 95C=100% | Keeps idle quiet, starts airflow before boost spike, avoids sudden jet noise |
| GPU Fan Curve (Thin Chassis) | 35C=0%, 50C=30%, 65C=55%, 75C=80%, 85C=100% | GPU is primary heat source in games; more aggressive mid-range than CPU |
| Hysteresis | 5C to 8C for both CPU and GPU | Prevents rapid on/off cycling when hovering at a threshold |
| Response Time / Ramp Delay | Up / Ramp Up: 1.5 sec, Down / Ramp Down: 4.0 to 6.0 sec | Slows audible pulsing during loading screens and menu stutters |
| Zero RPM / Fan Stop | Enabled below 40C to 45C, Disabled above 45C | Silence for browsing while ensuring fans spin before heat soak |
| Sensor Source | CPU: CPU Package, GPU: GPU Hotspot | Prevents fans from reacting to single edge core vs actual die heat |
| Max Temp Override | Force 100% at 95C CPU / 88C GPU Hotspot | Safety ceiling that overrides curve if paste or dust is degrading |
Setting-by-Setting
CPU Fan Curve Points: 30C=0%, 45C=25%, 60C=45%, 75C=70%, 85C=90%, 95C=100%
What causes it: Thin laptops boost the CPU to its power limit for seconds at a time, spiking from 45C to 75C instantly. A flat or late curve causes thermal throttling before fans respond, then fans jump to 100% and create an annoying cycle.
How to check: In HWiNFO64, enable Sensors Only and watch CPU Package temperature during a Cinebench R23 single run. In FanControl, go to Dashboard and observe which curve point your CPU is hitting. In G-Helper, press Fn + F5 and open Fans & Power to see the live graph.
What to do: Open FanControl and create a Custom Graph. Click Add Graph > Custom. Set sensor to CPU Package. Add six points and drag to the values above. Path in MSI Center: Features > User Scenario > Advanced > Fan Speed > Custom > click gear icon next to CPU Fan. Path in Armoury Crate: Device > System > GPU Power Saving Off > Fan Curve > Manual. Set the same six points using the pencil icon for each node. Set Fan Start value to 25% and Fan Stop to 0% with Stop temp at 45C.
How to undo: In FanControl, click Disable Control and toggle Manual Override off, or delete the Custom Graph. In MSI Center or Armoury Crate, select Auto or Balanced from the curve preset dropdown to restore default. Config file to revert in FanControl is located at C:\Users\YourName\AppData\Roaming\FanControl\fanControl.json – rename it to fanControl.json.bak and restart.
GPU Fan Curve Points: 35C=0%, 50C=30%, 65C=55%, 75C=80%, 85C=100%
What causes it: The GPU produces sustained heat under load, not brief spikes. A CPU-linked shared-fan curve will undercool the GPU in native GPU-bound games, causing hotspot throttling and chassis heat soak that then warms the CPU as well.
How to check: Run a game with an overlay like MSI Afterburner OSD (Toggle with assigned hotkey) and watch GPU Temperature and GPU Hotspot. If Hotspot exceeds edge temp by more than 15C and fans are still below 60%, your curve is too passive. In HWiNFO64 check GPU Fan RPM vs GPU Hotspot.
What to do: In FanControl, create a second Custom Graph with sensor set to GPU Hotspot, not GPU edge. Assign it to GPU fan header (usually Fan 2 on thin models). In G-Helper path: Fans & Power > Fan Curve Editor > Select GPU > Switch Sensor to GPU Hotspot in the dropdown. Enter points: 35C at 0%, 50C at 30%, 65C at 55%, 75C at 80%, 85C at 100%. If your laptop has a single shared heatpipe with one control, link both fans to the higher of CPU Package and GPU Hotspot using the MIX function: Add Control > Mix > Select both graphs > Function = Max.
How to undo: In FanControl, change the MIX control back to Auto or reassign fan to default profile. In G-Helper, click Apply > Default. In Armoury Crate, click Reset to Standard.
Hysteresis: 5C to 8C
What causes it: Without hysteresis, fans turn on at 60C and off at 59C in a loop, creating audible pulsing every few seconds when idling near the threshold.
How to check: Listen for rhythmic revving at idle and watch Fan RPM graph in FanControl. If RPM oscillates between 0 and 1500 every 5 to 10 seconds while temperature is stable within 2C, hysteresis is too low. Check setting at Configuration > Hysteresis in FanControl or the small H icon next to each point in G-Helper.
What to do: In FanControl, select your CPU Graph > Edit > Hysteresis > set to 6C. Set the same for GPU Graph. In MSI Center > Fan Speed > Custom > Advanced Settings > set Hysteresis to 7C. In BIOS on some thin models (Access with Del or F2 at boot > Advanced > Thermal Configuration > Smart Fan > Hysteresis), set to 5C if software control is not persistent after sleep.
How to undo: Set hysteresis back to 0C or 2C (default) in the same menu. Fan behavior will immediately return to rapid switching, which confirms the setting was active.
Response Time and Ramp Delay: Up 1.5 sec, Down 4.0 to 6.0 sec
What causes it: Instant ramp-up makes fans react to micro-spikes from background tasks. Instant ramp-down makes them drop and then immediately spin up again during game loading stutters.
How to check: Open a web browser with multiple tabs and watch for fan blips. In FanControl, look at the Response Time fields under each Fan Control card. Default is often 0.0 sec for both directions.
What to do: For each fan card in FanControl, set Response time Up to 1.5 seconds and Down to 5.0 seconds. In G-Helper, this is labeled Spin Up/Spin Down Time under Fans & Power > Advanced. In Armoury Crate, labeled Reaction Time. This delays the audible change without ignoring true heat buildup. The range remains 0.1 to 10.0 seconds; do not set Down longer than 6.0 on thin laptops or you risk holding low RPM while a new load starts.
How to undo: Set Up to 0.5 sec and Down to 0.5 sec. Changes apply instantly without a restart.
Zero RPM Mode and Acoustic Limit
What causes it: Zero RPM keeps fans off for silence, but if enabled too high it traps heat and causes a sudden wall of noise when they finally kick in at 70C. An acoustic limit caps max RPM to reduce noise but can cause throttling if set too low on a thin chassis.
How to check: In idle on battery, check if fans are at 0 RPM in HWiNFO64. Check Acoustic Limit in BIOS path Advanced > Acoustic Noise Settings or in MSI Center > User Scenario > Fan Speed > Noise Cancellation toggle. If CPU Package idles above 55C with fans at 0, the start threshold is too high.
What to do: Keep Zero RPM enabled but cap it: Set Fan Stop Temp 40C to 45C and Fan Start Temp 45C to 50C in FanControl (Edit graph > check Fan Stop). Set Fan Start PWM to 25% minimum, never 10%, because thin 12V fans stall below 20%. For Acoustic Limit, leave it off for AC gaming. Only enable on battery with a cap of 65% max fan speed via FanControl > Limits > Max Speed 65%.
How to undo: Uncheck Enable Fan Stop. Set Acoustic Limit to Disabled. Fan will idle at 1800 to 2200 RPM instead of 0.
By Hardware Tier
Low Tier: Intel Core i5-13420H / AMD Ryzen 5 7640U, RTX 4050 Laptop or RTX 3050, Single Fan or Shared Heatpipe
What causes it: Low-tier thin laptops use one fan and a single shared heatpipe, so CPU and GPU heat compete. Power limits are low (45W to 65W CPU), but cooling capacity is also minimal.
How to check: Look at vent count – one exhaust vs two. In HWiNFO64 check if CPU and GPU fans report as one RPM sensor. Check Device Manager > System devices for thermal zone count.
What to do: Use a combined curve with Mix set to Max as described above. Keep mid-range more conservative to avoid overshoot: 60C=50%, 70C=75%. Do not use Zero RPM above 40C on these models; set Stop at 40C, Start at 45C. In Windows Settings > System > Power > Power mode, set to Best Performance while on AC so power plan does not fight your curve. File location for manual backup before changing: Export profile in G-Helper to C:\ProgramData\GHelper\profiles.json.
How to undo: In G-Helper load Profiles > Default, or delete the Mix control and assign fan back to CPU sensor only.
Mid Tier: Intel Core i7-13650HX / AMD Ryzen 7 7840HS, RTX 4060 / 4070 Laptop, Dual Fan Separate Pipes
What causes it: Mid-tier thin chassis have dual fans but thin fin stacks. They can handle sustained 90W to 115W GPU load if airflow ramps early, otherwise the chassis soaks and keyboard becomes uncomfortable.
How to check: Confirm dual fan control in FanControl – you should see Fan1 and Fan2 controllable separately. Check Armoury Crate Device > Resource Monitor for independent fan RPM.
What to do: Apply the Quick Recommended Settings exactly as listed. For CPU, keep hysteresis at 5C. For GPU, set hysteresis to 7C because GPU temps are smoother. Set ramp down to 4.5 sec. If your laptop has a MUX Switch (Armoury Crate > System > MUX or BIOS > Advanced > Graphics Configuration > MSHybrid), set to Ultimate/Discrete GPU mode for gaming so GPU fan reacts to direct output, not Optimus overhead.
How to undo: Switch MUX back to Hybrid and set curves to Auto. Curves will revert after reboot if you do not click Save.
High Tier: Intel Core i9-14900HX / AMD Ryzen 9 7940HS, RTX 4080 Laptop, Vapor Chamber or Triple Pipe Thin Chassis
What causes it: High-tier thin models push 140W+ combined. Vapor chambers spread heat well but saturate quickly; they need higher sustained airflow to prevent sharp CPU package spikes from dragging down GPU clocks.
How to check: Check CPU Package Power (PPT) in HWiNFO64 during gaming. If it sustains above 65W with temps above 88C, you need more mid-range airflow. Check if BIOS has Performance/ Turbo modes that override custom curves – path BIOS > Advanced > Thermal Configuration > Performance Mode.
What to do: Raise mid-range slightly: CPU 60C=50%, 75C=75%, keep 95C=100% ceiling. Keep GPU as recommended but add a step at 70C=70% to stay ahead of hotspot. Set Up Response to 1.0 sec (faster) and Down to 6.0 sec. If your app has Fan Curve Sync, disable it and control CPU and GPU fans independently; synced control overcools one side and undercools the other in CPU-heavy vs GPU-heavy games. In MSI Center > Features > User Scenario > Extreme Performance, disable Auto Fan Boost if you use custom – they conflict.
How to undo: Re-enable Sync and set preset to Extreme Performance. In FanControl, set both controls to use Temperature Offset 0 and disable independent graphs.
Common Mistakes
Setting all points to a straight line from 0% to 100% creates loud idle and weak mid-range cooling. The curve should be flat at the bottom, moderate in the middle, steep at the top, not diagonal.
Using CPU Average instead of CPU Package as the sensor causes delayed response. Package reflects the hottest core instantly; average smooths spikes too much for a thin chassis.
Forgetting to disable manufacturer fan boost after setting a custom curve. Armoury Crate Turbo, MSI Center Cooler Boost (toggled with Fn + arrow key or button on supported models), and Lenovo Vantage FN+Q Performance Mode will override FanControl until disabled. Always set the manufacturer app to Manual/Custom first.
Setting minimum fan speed below 20% PWM. Thin 5V/12V fans stall and click below this, causing bearing wear and inaccurate RPM reporting. Keep minimum at 25% or 1400 RPM.
Saving curves only in software that does not load at startup. In FanControl, go to Settings > Start with Windows and check Autostart and Start minimized. In G-Helper, check Startup in Task Scheduler > GHelper. If you use BIOS Smart Fan only, changes persist without autostart but lack hysteresis control.
FAQ
Should I use BIOS or software for fan curves on a thin gaming laptop?
What causes it: BIOS control is persistent but coarse, often only three points and no per-app switching. Software gives more points and hysteresis but can fail to load after a crash.
How to check: Enter BIOS with Del or F2 at startup > Advanced > Hardware Monitor or Thermal Configuration. If you see only three temperature/speed pairs, it is limited.
What to do: For most users, use FanControl or G-Helper for daily control and keep BIOS on Default/Auto as fallback. Ensure software autostart is enabled at Settings > General > Run at Startup. Test by restarting and confirming FanControl tray icon loads before opening a game.
How to undo: Uninstall or disable autostart for FanControl, then reboot. BIOS control resumes automatically.
Why do fans pulse loudly in menus or loading screens?
What causes it: Menu screens often run unlocked frame rates, spiking GPU load briefly, and loading screens spike CPU decompression. Zero ramp delay makes fans react instantly to each micro-spike.
How to check: Cap frame rate off and watch GPU Hotspot in overlay during a menu. If temp jumps 10C for one second and fans follow, ramp delay is the cause.
What to do: Set GPU curve Response Time Down to 5.0 sec and Up to 1.5 sec, and enable a 60 FPS or 90 FPS cap in your manufacturer app or in-game VSync setting to reduce menu spikes. In FanControl, increase Hysteresis from 5C to 7C.
How to undo: Return Response Down to 0.5 sec and disable frame cap. Pulsing will return immediately.
Is it safe to run fans at 0 RPM on a thin laptop?
What causes it: Thin chassis have little passive cooling, so heat soaks into the chassis and SSD quickly when fans are off above 45C.
How to check: Idle for 10 minutes with fans at 0 RPM and monitor CPU Package and SSD temperature in HWiNFO64. If SSD exceeds 50C at idle or CPU Package stays above 60C, passive cooling is insufficient.
What to do: Allow 0 RPM only below 40C to 45C. In FanControl graph, set Fan Stop at 42C and Fan Start at 48C with 25% PWM. For video playback or light work on a desk, this keeps silence without soak. On a bed or soft surface, disable Zero RPM entirely.
How to undo: Uncheck Fan Stop in the graph editor and set minimum to 25% at 30C.
Do I need different curves when on battery?
What causes it: On battery, power limits are lower, fan noise is more noticeable, and battery wear increases with heat. Using the AC gaming curve on battery wastes charge and creates unnecessary noise.
How to check: Go to Windows Settings > System > Power & battery > Power mode. Check Battery Saver threshold and check if your manufacturer app switches profiles automatically (MSI Center Profile Switch, G-Helper Auto Profiles based on power source).
What to do: Create a second profile named Battery Quiet. In FanControl, duplicate your graphs and reduce all points by 15%: example 60C=30% instead of 45%. Enable Acoustic Limit of 65% in G-Helper > Power Options > On Battery > Silent. Assign auto-switch: G-Helper > Settings > Auto profiles > On Battery AC Detached > Load Battery Quiet.
How to undo: Set power profile to Balanced and disable Auto profiles. Fan will use AC curve on battery until changed.
Related guides
- Best Vrr Settings for Gaming Laptop Display: Best Settings and Config (2026): 8 Tested (2026)
- Best Display Color Settings for Video Editing Laptop: Best Settings and Config (2026): 8 Tested (2026)
- Best Windows Power Settings for Creator Laptop: Best Settings and Config (2026): 8 Tested (2026)
- Best Settings for Video Editing Laptop Performance: Best Settings and Config (2026): 8 Tested (2026)
