CPU Burn-in: How to Stress-Test Your Processor Safely

Quick CPU Burn-in Checklist for Overclockers and System Builders

1. Prepare hardware

• Clean cooling: Re-seat cooler and replace thermal paste if older than 1 year.
• Stable power: Use a reliable PSU with adequate wattage and quality cables.
• Good airflow: Ensure case fans are oriented for proper intake/exhaust.
• Known-good RAM & storage: Use tested modules and avoid mixing kits.

2. Update firmware & drivers

• BIOS/UEFI: Update to latest stable release for improved stability and CPU microcode.
• Chipset drivers: Install latest chipset/board drivers from manufacturer.

3. Configure baseline settings

• Load optimized defaults: Start from stock or vendor “optimized” profile.
• Disable auto features temporarily: Turn off XMP/DOCP, power-saving C-states, and CPU boost if isolating stability.
• Set conservative voltages: Use safe voltage ranges for initial runs (follow CPU/vendor guidance).

4. Monitoring tools

• Temperatures: HWInfo, CoreTemp, or Ryzen Master/Intel XTU.
• Voltages & clocks: HWInfo or BIOS logging.
• System logs: Windows Event Viewer or motherboard debug LEDs/post codes.

5. Stress-test sequence (ordered)

1. Light stability: Run Prime95 Small FFTs or Linpack for 15–30 minutes to verify basic operation.
2. Sustained load & thermal check: Run AIDA64 FPU or Prime95 with AVX disabled for 1–2 hours while monitoring temps.
3. AVX stress (optional): If using AVX workloads, run with AVX enabled for 30–60 minutes at target clocks—expect higher temps/power.
4. Memory & platform: Run MemTest86 (bootable) or MemTest in OS for several passes.
5. System burn-in: Run a combined workload (e.g., Cinebench loop + Unigine/3DMark) for 2–4 hours to test real-world stability.

6. Pass/fail criteria

• Pass: No crashes, no thermal throttling, stable voltages, and temperatures within safe limits for the planned use.
• Fail: Crashes, blue screens, unexplained reboots, or temps exceeding manufacturer Tmax — stop immediately and troubleshoot.

7. Troubleshooting steps

• If overheating: Re-check mounting, reapply thermal paste, increase fan curve, consider better cooler.
• If instability: Raise voltage in small increments (e.g., +0.01–0.05 V) or back off clock/overclock.
• If RAM errors: Lower frequency or loosen timings, or test modules individually.
• If unexplained reboots: Check PSU, power cables, and motherboard VRM temps.

8. Finalize settings

• Re-enable desired features (XMP/DOCP, boost) and perform a shorter verification run (1–2 hours).
• Keep profiles saved in BIOS and