Benchmarking on XTU give 15 degrees higher temps than stress tests

[mcgge1360]

...why? Stress testing on XTU as well as CPU mining (as a stress test, not for money) will reach a max of ~55. If I benchmark it goes to 69. Why does this happen, and should I use benchmark or stress test as an indicator for temperature when overclocking?

i3-8350K
z-370 asrock pro4

 

[zebarjadi.raouf]

...why? Stress testing on XTU as well as CPU mining (as a stress test, not for money) will reach a max of ~55. If I benchmark it goes to 69. Why does this happen, and should I use benchmark or stress test as an indicator for temperature when overclocking?

i3-8350K
z-370 asrock pro4

The benchmark might use different instructions like AVX which outputs much more heat to give a score. Stress test mainly stresses the CPU for stability purposes.

It's recommended to use Prime95 v26.6 to stress test/find reasonable max. If you want the max temperatures possible, get the latest Prime95 which has AVX support.

 

[mcgge1360]

...why? Stress testing on XTU as well as CPU mining (as a stress test, not for money) will reach a max of ~55. If I benchmark it goes to 69. Why does this happen, and should I use benchmark or stress test as an indicator for temperature when overclocking?

i3-8350K
z-370 asrock pro4

The benchmark might use different instructions like AVX which outputs much more heat to give a score. Stress test mainly stresses the CPU for stability purposes.

It's recommended to use Prime95 v26.6 to stress test/find reasonable max. If you want the max temperatures possible, get the latest Prime95 which has AVX support.

I've been reading a bit about what you said and it seems AVX is overkill for stress testing unless you actually do AVX workloads (I don't). Should I use the built in XTU stress test or prime95?

 

[zebarjadi.raouf]

Should I use the built in XTU stress test or prime95?

I prefer Prime95 + RealBench. You could use XTU as well. Any decent stress tester should get the job done.

 

[CompuTronix]

...why? Stress testing on XTU as well as CPU mining (as a stress test, not for money) will reach a max of ~55. If I benchmark it goes to 69. Why does this happen, and should I use benchmark or stress test as an indicator for temperature when overclocking?

i3-8350K
z-370 asrock pro4

If you're trying to determine your rig's thermal performance, then you have to take a closer look at your approach. Methodology is the key.

“Stress” tests vary widely. Gaming, applications, rendering, transcoding and streaming are partial, fluctuating workloads with fluctuating temperatures, which aren’t well suited for testing thermal performance. XTU, mining and benchmarking are also inappropriate, because they all run at different load levels other than 100%.

Intel tests their processors at a steady-state 100% TDP workload to validate thermal performance. Although Intel's software is proprietary, there is one freeware utility which replicates Intel's test methodology; Prime95 version 26.6 Small FFT's. Since you don't use AVX, do NOT use later versions. Run only Small FFT's for just 10 minutes.

• Prime95 v26.6 - http://www.mersenneforum.org/showthread.php?t=15504

Utilities that don't overload or underload your processor will give you a valid thermal baseline. Here’s a comparison of utilities grouped as thermal and stability tests according to % of TDP, averaged across six processor Generations at stock settings rounded to the nearest 5%:

Higher TDP tests produce higher Core temperatures. All tests will show 100% CPU Utilization in Windows Task Manager, which indicates processor resource activity, not % TDP workload. Although actual Power dissipation (Watts) varies with Core Speed, Core voltage and workload, Prime95 v26.6 Small FFT’s always provides a steady 100% workload, whether you’re running stock or overlocked.

It's all in here: Intel Temperature Guide - http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html

Give it a read.

CT

 

[william p]

Benchmarking will stress the CPU and GPU at the same time. This will produce more total heat inside the case. Stress testing usually stresses just one or the other. Case air flow might be the issue.Either more cold air in, or more hot air out. Many GPUs dump the hot air randomly inside the case, others force it out the back slots. Some send 1/2 of it forward where it may recirculate into the CPU cooling.

 

[zebarjadi.raouf]

Benchmarking will stress the CPU and GPU at the same time. This will produce more total heat inside the case. Stress testing usually stresses just one or the other. Case air flow might be the issue.Either more cold air in, or more hot air out. Many GPUs dump the hot air randomly inside the case, others force it out the back slots. Some send 1/2 of it forward where it may recirculate into the CPU cooling.

Depends on the benchmark. XTU should be pure CPU. If it was a gaming benchmark or userbench, then all the components would have been involved.