Recent content by Martin

Many modern systems support changing memory frequency on the fly depending on load.

Yes, the polling rate affects it but it will be still too large for such a long period. Try to disable monitoring of the PresentMon sensor if that will solve the problem.

Sorry, this isn't currently possible neither in the Pro version.

Probably same problem as this: https://www.hwinfo.com/forum/threads/hwinfo64-cpu-die-average-spike.8044/

Yes, normal.

PCI(e) bus has a different topology. You can examine this in the main HWiNFO window.

HWiNFO shows GPUs in the order in which they appear on PCIe bus.

Not sure what exactly do you mean without seeing a screenshot but if you have a GIGABYTE board then it's most likely the AORUS theme used by HWiNFO. This can easily be disabled in main settings.

That's intentional for these aggregated values.

Expand the nodes by clicking the arrow to the left.

Close HWiNFO. Run "regedit" tool Navigate to: HKEY_CURRENT_USER\Software\HWiNFO64\Sensors\GpuI2cCache Delete the entire key. If that won't help, please attach the HWiNFO Debug File.

Do that reset right after starting HWiNFO (before the scan is started). Or restart HWiNFO right after the reset.

Sorry, there are no dedicated shortcuts/hotkeys for those features.

Try to "Reset GPU I2C Cache" in HWiNFO settings - "SMBus / I2C" section.

If voltage offsets are not shown, then most likely because Windows HVCI (Core isolation/Memory integrity) is blocking access and you will need to disable it to see them.