Undervolting: lower temperature 10°C no new cooler

Does the laptop often freeze when work is important? Does the PC become a jet engine as soon as the game opens? Don’t rush to buy an expensive cooler. There is a smart solution called UNDERVOLTING that can make your hardware cooler, more power-efficient, and faster!

Have you ever been annoyed by a laptop or PC suddenly getting very hot, a noisy fan like a plane taking off, or a battery draining quickly when only used for light tasks? This problem is often caused by an excessive voltage being sent to the processor, making it work harder than it should.

This is where undervolting comes in as a smart solution, a technique that allows you to reduce the voltage on your CPU or GPU without sacrificing performance. Result? Hardware that is cooler, more power-efficient, and has the potential to be more durable.

Imagine being able to lower the temperature of the processor by 5-10°C with just a few software adjustments, or make the laptop battery last longer without the need for hardware upgrades. For example, a gaming laptop with an NVIDIA GPU can experience a significant temperature drop from 85°C to 72°C after being undervolted, while gaming performance remains stable. However, this technique should not be done carelessly, misconfigurations can cause the system to be unstable or even crash.

Benefits of Undervolting: More Than Just Power Savings

Undervolting is not only a strategy to reduce electrical power consumption but also an effective solution to improve the stability, efficiency, and durability of hardware in the long run. Here are some of the key benefits you can feel:

1. Cooler Temperature, More Stable Performance

An overheating processor triggers thermal throttling, an automatic mechanism that reduces clock speed to prevent damage. Consequently? Games lag, video renders longer, or apps suddenly freeze.

2. Laptop Battery Lasts Longer

Undervolting optimizes power efficiency by reducing voltage “waste”. In the laptop ecosystem, power efficiency has a huge impact on battery life. When the CPU/GPU does not receive overvoltage, the power consumed automatically becomes lower.

3. Longer Component Life

Heat is the main enemy of electronics! Any drop of 10°C can double the lifespan of the transistor (based on the Arrhenius Equation).

Each electronic component has a heat and voltage threshold. Continuous exposure to heat can speed up the silicon degradation process, shortening the lifespan of components such as CPUs, GPUs, and VRM (Voltage Regulator Module).

“Undervolting is a long-term investment. In addition to maintaining temperature, this technique reduces stress on components, similar to a car engine that is not constantly forced to work at high RPMs.”

The Risks of Undervolting (And How to Avoid It)

Although undervolting is safe for hardware, it does not mean that this technique is risk-free. Misconfigurations, especially if done carelessly, can cause serious problems. Here are the main risks that may occur along with practical solutions to avoid them:

1. System Crash & Blue Screen (BSOD)

The voltage is too low, so the processor does not get enough power for stable operation. Reducing the offset voltage of Intel’s CPU to -150 mV can cause a boot loop or fail to enter Windows.

Solution:

• Gradually lower the voltage (e.g., -10 mV per experiment).
• Use ThrottleStop (CPU) or MSI Afterburner (GPU) to monitor stability.

2. Application & Game Instability (Freeze/Artifacts)

Extreme undervolting can also trigger graphics artifacts, sudden freezes, or self-exiting applications (crashes), especially when components are working under high loads. The GPU/CPU is unstable when working at high loads, resulting in:

• Screen tearing or visual artifacts (on GPUs).
• The application suddenly becomes unresponsive (especially when rendering/video editing).

Solution:

• Perform a stress test after each change:
  • CPU: Use Prime95 or Cinebench.
  • GPU: Use FurMark or 3DMark Time Spy.
• Monitor temperature & clock speed with HWMonitor or GPU-Z.
• If an error occurs, increase the voltage slightly until it stabilizes.

3. Performance Drops (Excessive Undervolt)

Some processors (such as Intel 12th/13th Gen) use a voltage-dependent adaptive boost algorithm. If the undervolt is too extreme, the CPU instead reduces the clock speed to compensate.

How to Detect:

• Compare benchmark scores (before & after undervolt) on Cinebench R23.

“Undervolting is like driving on a slippery road, it requires fine adjustments, not just full throttle!”

Undervolting Guide for Beginners: Secure Step by Step

For beginners, the Undervolting process can seem technical and confusing. But with the right approach and systematic measures, you can do it safely to improve system efficiency without sacrificing performance.

Step 1: Advance Preparation (Mandatory!)

Before undervolting, make sure you have properly set up your tools and systems to prevent the risk of damage or data loss.

Required Software:

Preventive Measures:

1. Backup important data (anticipation of blue screen)
2. Record the default voltage and clock speed values
3. Make sure the BIOS is up-to-date (especially for laptops)
4. Do it in a stable working environment (do not undervolt when the laptop is in battery mode)

Step 2: Undervolting Steps

For Intel processors:

1. Open ThrottleStop
2. Click “FIVR”
3. In the “CPU Core Voltage” section:
   • Check “Unlock Adjustable Voltage”
   • Set “Offset Voltage” to -0.050V (50mV)
   • Click “Apply”

For AMD Ryzen Processors:

1. Open Ryzen Master
2. Select “Curve Optimizer”
3. Arrange:
   • All Core Negative Offset
   • Starting from -10
   • Test the stability of each 5-point drop

For GPUs (NVIDIA/AMD):

1. Open MSI Afterburner
2. Press Ctrl+F to open the curve editor
3. “Flatten Curve” Technique:
   • Select 900mV point
   • Pull the clock speed to the desired value
   • Flatten the curve after that point

Step 3: Start Undervolting Gradually

1. Lower 10 – 15 mV per experiment (don’t go straight to -100 mV!)
2. Save the profile of every change
3. Example of Initial Setting:
   • Intel CPU: -50mV offset
   • NVIDIA GPUs: Lower 50MHz every 25 mV

Step 4: Stress Test (Key to Success!)

After any voltage reduction, it is very important to test the stability of your system. This is to ensure that the CPU or GPU can still perform at its maximum without experiencing crashes or throttling.

Stability Testing Tools:

Success Indicators:

• The system did not experience a BSOD or a restart during testing
• No artifacts
• The benchmark score does not drop drastically compared to the default voltage
• Lower temperature 5-10°C but clock speed remains stable