XTU evaluates undervolting techniques
XTU evaluates undervolting techniques
I have been attempting to lower my CPU voltage by -0.110mV in order to cut down the unusually high temperatures in my laptop (Inspiron 7567, i7 7700HQ core, 1050ti, 16GB RAM). I lack any background in under or overclocking. This adjustment has lowered the temperature, but I’m facing difficulties in checking for stability and finding the various tests on XTU quite confusing. I searched online, but the information was hard to understand. If someone could clarify this for me, it would be very helpful.
The XTU stress test includes four sections: CPU stress test, CPU stress test with AVX, CPU stress test with AVX2, and memory stress test. I’m unsure how these differ from each other and don’t fully grasp what AVX stands for. Could you explain which tests are recommended and why?
I performed each of these tests for 10 minutes while the laptop was powered on; all results were successful. I also ran the DELL stress test, which generates higher temperatures, and it passed as well—this made me think the system was stable. Should I have extended the testing time?
During the night, I left the laptop running but unplugged. When I woke up, it restarted during the night (returning to the login screen), and the reliability monitor reported "Windows was not properly shut down" (miscellaneous failure). There were no BSODs recorded, which I expected if there wasn’t enough voltage. Could it be that the laptop can tolerate undervolting while plugged in but not when powered off? Or could the restart have triggered an issue?
Most applications don't rely on AVX or AVX2, so typically you run tests with a non-AVX workload. AVX testing generates significantly more heat and may not accurately reflect real-world performance.
You usually perform AVX/AVX2 tests during overclocking to prevent overheating. Certain settings (AVX offset) let you lower the CPU frequency when it's handling those tasks.