Skip to main content

How to Stress-Test CPUs and PCs (Like We Do)

Conclusion: The Best Stress-Test Tools & Settings

Testing Maximum CPU Load

If your goal is to maximize power consumption and generate as much waste heat as possible to ensure your thermal solution is sufficient, then power viruses like Prime95 with AVX and Small FFTs, powerMAX, or AIDA64’s Stress FPU option are the way to go. However, these stress tests aren’t representative of the loads from commonly used applications. That's because most desktop software doesn't use AVX instructions the way artificially demanding workloads do. To simulate everyday scenarios, stress tests that tax your hardware using SSE code paths are the way to go.

Combine a tool like that with MSI Kombustor’s GPU Core Burner or Memory Burner for an excellent worst-case platform scenario. Even a game like The Witcher 3 can help you create similar or worse power consumption numbers than a lot of popular graphics stress tests.

Optimizing CPU Cooling

Your cooler should be set up so that it’s able to deal with these extreme loads, regardless of the code path responsible for them. And don't forget chassis ventilation. Many enthusiasts build great-looking custom water-cooling loops, but then forget that the motherboard needs active cooling in certain areas, else it'll overheat as well. Without at least some airflow, the board's VRMs might not last long.

Optimizing the Noise Level

Fortunately, airflow through your chassis can be temperature-controlled. Once the maximum temperature is known, a good combination of CPU and GPU stress testing can be used to simulate demanding gaming loads. MSI Kombustor and OCCT are particularly well-suited for this task if they’re configured as we described. The results of those tests may then be used to form the basis for fan-curve optimization by reserving the most aggressive fan speeds for the range approaching your ceiling.

Stability Testing & Its Limits

None of the pure stress tests guarantees 100% stable everyday operation. All they’re really good for is finding the system’s thermal and electrical limits. Everything else, especially when it comes to overclocked components, can be truly found out only by using suitable applications run over a long period of time. Prime95 and its contemporaries aren’t really designed to find errors, because they are fairly simple computational tools that serve one purpose: to consume a lot of power. (At least that's the case for Prime95's torture-test mode.)


With some thought and the right applications, most of which are free, you're able to test for cooling capacity and operational safety. These tests can (and should) be repeated periodically in order to suss out any changes due to aging components. In the beginning, testing every six months is sufficient, with the intervals between tests decreasing as the PC ages, especially if it’s used a lot or very intensively. It doesn’t matter if it’s been overclocked or not. Better safe than sorry!

MORE: Best CPUs For Gaming 

MORE: Intel and AMD Processor Benchmark Hierarchy Comparisons

MORE: All CPUs Content