Problem: NVIDIA GeForce Driver 595.59 (Feb 2026) breaks GPU fan control/monitoring, risking overheating and instability
Published: 2026-02-28 12:20 (local time)
Quick Summary
- A newly released NVIDIA GeForce driver (595.59) was widely reported to break GPU fan monitoring/control for some users, and NVIDIA temporarily pulled the downloads while investigating.
- Symptoms range from “only one fan detected” to fans not ramping properly after fan-stop, plus other reports like clock/voltage oddities and HDR issues.
- This is especially dangerous because the PC can look “fine” until the GPU overheats (throttling, black screens, crashes, forced restarts).
- There is no single universal fix besides rolling back to a previous driver (NVIDIA specifically pointed to 591.86 WHQL for affected users).
- Workarounds below focus on: verifying fans are physically spinning, rolling back safely, and reducing risk until a corrected driver/hotfix arrives.
What’s happening
In late February 2026, NVIDIA’s GeForce Game Ready/Studio driver version 595.59 triggered widespread user reports of broken GPU fan control and missing fan sensors (for example, only one fan showing up in monitoring tools, custom fan curves being ignored, or fans not properly resuming after fan-stop). NVIDIA acknowledged a bug and removed the driver downloads temporarily while it investigates, advising impacted users to roll back to driver 591.86 WHQL.
Reports are concentrated among GeForce RTX 50-series owners (where multi-fan cards are common), but coverage and user discussion indicate it may also affect other RTX generations depending on board partner implementations and monitoring stacks. The problem matters to gamers because it can turn a normally safe “install the latest driver” routine into an overheating risk during long sessions—especially in demanding new releases or betas that prompted the driver update in the first place.
Likely causes (what research suggests)
- Driver-level regression in fan telemetry/control paths: NVIDIA’s own acknowledgement and rollback guidance strongly suggest the core issue is inside 595.59 rather than a single third-party tool conflict.
- Interaction with fan-stop behavior and custom curves: Multiple user reports describe fans working at boot but failing to behave correctly once Windows loads and the driver takes over, or after fan-stop events.
- Monitoring/OC utilities amplify symptoms (not necessarily the root cause): Users frequently mention MSI Afterburner, Fan Control, HWiNFO, and vendor utilities. These may worsen or reveal the problem, but reports also exist without heavy tweaking, so treat them as “complicating factors,” not the single culprit.
- Additional regressions in the same driver build: Community and media coverage also mention voltage/clock caps (e.g., ~0.95V) and HDR issues; these may be separate bugs, but they support the idea of a broader stability regression in 595.59.
Solutions & Workarounds
1) Roll back to NVIDIA 591.86 WHQL (recommended baseline)
Who it helps: Any PC gamer who installed 595.59 and now sees abnormal fan behavior, missing fan sensors, overheating, crashes, or black screens.
Steps:
- Open the NVIDIA App.
- Go to the Drivers tab.
- Use the menu (commonly shown as three dots) and choose the option to reinstall/restore the previous driver.
- Reboot.
- After reboot, confirm fan behavior under load (see Solution #4).
Risks/tradeoffs: You may lose game-specific optimizations that were bundled with 595.59. Some titles/betas might recommend the newest driver, but stability and safe cooling matter more.
Stop & contact official support when: Fans still won’t ramp correctly after rollback, temps spike abnormally fast, or the system crashes under light GPU load.
2) Do a “clean” driver reinstall using DDU (stronger rollback when sensors stay broken)
Who it helps: Users who rolled back but still can’t regain fan sensors or proper fan curve control, or who have layered driver installs over time.
Steps:
- Download DDU from a reputable source (prefer official/major vendor guidance).
- Reboot Windows into Safe Mode.
- Run DDU and select GPU cleanup, then choose Clean and restart.
- Back in normal Windows, install a stable NVIDIA driver (such as 591.86 WHQL).
- Reboot and validate fan behavior and temperatures.
Risks/tradeoffs: Incorrect use can remove drivers you didn’t intend to remove; you’ll need to reinstall display drivers after. If you rely on special CUDA/Studio setups, plan your reinstall carefully.
Stop & contact official support when: Your GPU fans do not spin reliably even at idle/boot, or Windows becomes unstable after repeated driver reinstalls.
3) Remove fan/overlay/OC control layers temporarily (Afterburner, Fan Control, vendor GPU tools)
Who it helps: Players who must remain on a newer driver for a specific game/test (or troubleshooting) and suspect a tool conflict is making fan control worse.
Steps:
- Close and disable auto-start for MSI Afterburner / Fan Control / vendor tuning apps.
- Reboot.
- In NVIDIA settings or default GPU behavior, ensure the card is set back to “automatic” fan control (no custom curve).
- Re-test temperatures in a controlled load for 5–10 minutes (see Solution #4).
Risks/tradeoffs: You lose your preferred fan curve and undervolt/OC profiles. This is a diagnostic step, not a permanent fix.
Stop & contact official support when: Fans still fail with all third-party utilities removed (suggesting the driver/firmware path is at fault).
4) Validate cooling immediately (don’t trust “it feels quiet”)
Who it helps: Everyone affected or unsure—this reduces the chance of accidental overheating.
Steps:
- Open a trusted hardware monitor (HWiNFO or GPU-Z are common choices) and watch GPU temperature and fan RPM.
- Run a short, controlled GPU load (a game menu with unlocked FPS, a built-in benchmark, or a light stress test).
- Confirm: temperature rises gradually and fans ramp; RPM readings should be plausible and stable.
- If temperature rockets upward while fans remain at 0/very low RPM, stop the test immediately and proceed to rollback/clean reinstall.
Risks/tradeoffs: Any stress test can worsen overheating if fans truly aren’t responding—keep it short and monitored.
Stop & contact official support when: Fans never ramp physically (not just “no sensor”), or the card hits unsafe temps quickly.
5) Temporary risk-reduction if you can’t roll back right away: cap FPS and lower power
Who it helps: Players who are mid-event/tournament/session and need a short-term mitigation while preparing a rollback.
Steps:
- Enable an in-game FPS cap (e.g., 60–120) and disable uncapped menus.
- Lower graphics settings that spike power draw (ray tracing, ultra shadows, high-resolution upscalers at max quality).
- If you use a power limit slider (vendor tool), reduce the power target modestly.
- Monitor temps constantly; treat this as a short bridge to a rollback, not a “solution.”
Risks/tradeoffs: Reduced performance/latency characteristics; still not safe if fans truly fail to spin.
Stop & contact official support when: You cannot keep temps under control even with reduced load.
Prevention (so it doesn’t come back)
- Stay one stable driver behind when everything is working—especially on brand-new GPU generations.
- After any driver update, do a 2-minute “sanity check”: verify fan RPM changes and temps under a small load.
- Avoid stacking multiple tuning tools (pick one) and document your “known good” driver version.
- Keep a rollback plan: store the last stable driver installer and know how to enter Safe Mode for DDU if needed.
FAQ
Q: How do I know if this is affecting me?
A: Warning signs include missing fan sensors (only one fan shows), fans not ramping after fan-stop, unusually high temps, sudden throttling, black screens, or crashes soon after installing 595.59.
Q: NVIDIA pulled the download—does that mean the issue is confirmed?
A: NVIDIA acknowledged a bug and removed downloads temporarily while investigating, and advised rolling back if experiencing fan control issues.
Q: Is it safe to keep gaming if my card “seems fine”?
A: Not without monitoring. A broken fan curve can look “quiet” until temps climb; validate RPM and temperatures first.
Q: I rolled back but fan control is still weird—what next?
A: Do a clean uninstall using DDU in Safe Mode, then reinstall a known-good driver. Also remove third-party fan/OC utilities during testing.
Q: Does this affect only RTX 50-series?
A: Reports heavily feature RTX 50-series, but discussion and coverage suggest broader impact is possible depending on the card and tooling. Treat it as potentially affecting any NVIDIA user who installed 595.59.
Q: Should I wait for a hotfix driver?
A: If you’re already impacted, the safest path is rollback now; then you can update later once NVIDIA releases a corrected build and reports stabilize.