Comparing benchmarks on XTU shows temperatures 15 degrees above stress tests
Comparing benchmarks on XTU shows temperatures 15 degrees above stress tests
...why? Stress testing on XTU along with CPU mining will hit a maximum around ~55. If benchmarking it reaches 69, this is normal. The reason behind this variation and whether to rely on benchmark or stress test for temperature during overclocking remains unclear. i3-8350K z-370 asrock pro4
mcgge1360 :
...why? Stress testing on XTU along with CPU mining (as a stress test, not for profit) will hit a peak around 55. If benchmarking it reaches 69, what causes this difference? Should I rely on benchmark or stress test when monitoring temperatures during overclocking?
i3-8350K
z-370 asrock pro4
The benchmark could involve different instructions like AVX, which generate more heat and affect the score. Stress testing focuses mainly on CPU stability. It's advised to use Prime95 v26.6 for stress testing or finding a reasonable maximum. For higher possible temperatures, opt for the latest Prime95 versions with AVX support.
mcgge1360 :
...why? Stress testing on XTU along with CPU mining (as a stress test, not for profit) will hit a peak around 55. If benchmarking it reaches 69, what causes this difference? Should I rely on benchmark or stress test when monitoring temperatures during overclocking?
i3-8350K
z-370 asrock pro4
The benchmark could involve different instructions like AVX, which generate more heat and affect the score. Stress testing focuses mainly on CPU stability. It's advised to use Prime95 v26.6 for stress testing or finding a reasonable maximum. For achieving the highest possible temperatures, opt for the latest Prime95 versions with AVX support.
Zebarjadi.raouf :
mcgge1360 :
Understanding the issue involves checking why stress testing on XTU and CPU mining reaches a high temperature. The benchmark could be affected by different instructions like AVX, which increases heat output. Stress tests focus on stability rather than just performance. It's advised to use Prime95 v26.6 for stress testing or the latest Prime95 with AVX support if you aim for maximum temperatures. The built-in XTU stress test might not be sufficient unless you specifically run AVX workloads.
I favor Prime95 plus RealBench. XTU is an option too. Any good stress tester should suffice.
mcgge1360 :
...why? Stress testing on XTU along with CPU mining (as a stress test, not for profit) will hit a maximum around 55. If benchmarking it reaches 69. What causes this difference, and should I rely on benchmark or stress tests when adjusting temperatures during overclocking?
i3-8350K
z-370 asrock pro4
To assess your system's thermal behavior, you need to examine your setup more closely. The approach matters most.
Stress tests differ greatly. Activities like gaming, apps, rendering, transcoding, and streaming involve variable workloads with changing temperatures—conditions not ideal for evaluating cooling performance. Stress testing on XTU, mining, or benchmarking doesn’t match the 100% load these systems typically experience. Intel evaluates thermal efficiency by running processors at full TDP capacity under steady-state conditions to confirm performance. While Intel’s tools are exclusive, a free alternative called Prime95 version 26.6 exists; see the link for details.
Small FFT's. Since your setup doesn’t use AVX instructions, avoid newer versions. Limit testing to Small FFT’s for just ten minutes.
• Prime95 v26.6 – http://www.mersenneforum.org/showthread.php?t=15504
Tools that don’t overload or underload the CPU will offer a reliable thermal baseline. A comparison of utilities categorized by thermal and stability tests based on TDP percentage, averaged across six generations at default settings (rounded to 5%), shows: higher TDP tests yield elevated core temperatures. All assessments display 100% CPU usage in Windows Task Manager, reflecting processor activity rather than actual % TDP load. Prime95 v26.6 Small FFT’s consistently delivers a stable 100% workload, whether stock or overclocked.
The key points are clear: Intel provides guidelines for interpreting these results.
Benchmarking will simultaneously challenge the CPU and GPU, generating greater overall heat within the enclosure. Stress tests typically focus on one component at a time. The airflow inside the case could be the problem—either too much cool air entering or too much warm air escaping. Some GPUs expel hot air unpredictably inside, while others push it out through back slots. Others direct half of it forward, which might then loop back into the CPU cooling system.
William P explains that benchmarking increases pressure on both the CPU and GPU simultaneously, leading to higher internal heat generation. Stress tests typically focus on one component at a time. The airflow inside the case could be the problem—either too much cool air entering or too much hot air escaping. Some GPUs expel hot air unpredictably, while others push it out through back slots. A portion of the heat may travel forward, potentially re-entering the CPU cooling system. The exact issue depends on the specific benchmark being used.