Hi AJK,

This most likely means that the CPU is affected by energy-saving settings. You may experience a higher load than on another computer because, for example, the computer may be running cooler or has a slightly lighter load. As a result, the energy-saving settings (such as the Windows Power Plan) may decide to throttle the CPU to reduce power consumption. Such throttling means that the 100% on the CPU meter in Aximmetry actually represents much less performance than would otherwise be possible.
We recommend using Windows on the highest possible Power Plan setting.

Warmest regards,

Hello Eifert, 

Thank you for your answer, I setup both PCs on the High-Performance Power Plan , but i will check if the CPU in the Driver or the AMD software has maybe a power efficiency plan. Also the thermals are on par and didn't show any signs of overheating in AIDA64. 
I doubt this might come from this since the benchmarks are close to what we had at startup. 

Can there be loops in setups or such in Aximmetry that would cause this kind of issues ?

If you need captions / screenshots of any kind feel free to ask. 

Thank you again for your help, 

Best regards, 

AK

Hi,

There’s nothing in Aximmetry that would cause this on its own.

However, I can’t say the same about Unreal. Your compound includes an Unreal module, and its processing is also reflected in Aximmetry’s Processor Load panel.

As a test, try running [Tutorials]\Instancing\Instancing - 03 - Script.xcomp, which contains no Unreal modules but can still put a significant load on the CPU. You will likely need to increase the number of instanced objects. You can do this by multiplying the Count pin value in each Random module by 5 (or another factor).

Also, you could check CPU usage in Windows Task Manager (Ctrl + Shift + Esc) on each computer and compare the results. Set the CPU graph view to Logical processors, which shows per-core usage:

Note: Aximmetry’s Processor Load panel uses a different calculation method than Windows Task Manager. As a result, it effectively reflects the utilization of the most heavily loaded CPU core / logical processor, rather than an overall average across all cores.

Warmest regards,