Squall Leonhart

Does HwInfo show them?

https://techcommunity.microsoft.com/blog/windowsservernewsandbestpractices/announcing-native-nvme-in-windows-server-2025-ushering-in-a-new-era-of-storage-p/4477353 NVME's will no longer be scsiport devices when this spec is enabled widely. @Fiery Home/Pro can turn it on with reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Policies\Microsoft\FeatureManagement\Overrides /v 1853569164 /t REG_DWORD /d 1 /f reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Policies\Microsoft\FeatureManagement\Overrides /v 156965516 /t REG_DWORD /d 1 /f reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Policies\Microsoft\FeatureManagement\Overrides /v 735209102 /t REG_DWORD /d 1 /f

these bogus readings appear on Intel z690+ too, when EC reading is on.

when closing aida64?

https://github.com/ysc3839/win32-darkmode/tree/delayload Technically it works on 10 as well, but there are some things to be concerned about, such as scroll bars, and the menu bar.

it doesn't stay closed because you turned on a plugin in an external application, such as MSI Afterburner which automatically starts aida64 so it can receive data.

update the asus ai suite software or remove it.

These are supplied by Nvidia NVAPI, AMD has no equivalent api.

you may have a HDD that has a high latency for SMART polling.

the cpu usage on the Performance tab is not consistent with utitilities or the usage shown on the Details tab, this is a known issue where the performance and processes tab takes boost clocks and core count awake and used into account when factoring usage, because it factors these into things, the usage on these tabs are usually higher than the Details tab. For the clock reading on task manager and aida64, you need a power profile that has the min cpu state set under 100% to report correct idle behavior having min cpu state in the power profile maxed out to 100 will show constant max clocks in aida64, task manager will show up to 200mhz range from max clock HWInfo will show max actual clock, and low effective clock.

because it was win11s fault.

Fix is in 22621.900 (Optional Preview) 22621.1020/22623.1020 (Beta)

Microsoft has resolved this on their end, aida64 cpu usage without any fixes matches MSIAB with the "fix"

It looks correct on this end too, 22621.1020

probably a stuck register in the CPU, fiery needs an ISA dump i believe.

Its just a matter of snapshot interval, Aida64 is pulling the info from the nvidia api.

the older driver also resolves a resume from hibernate freeze that was occuring

NtQuerySystemInformation(SystemProcessorIdleInformation) can be used to get the right values, its only NtQuerySystemInformation(SystemProcessorPerformanceInformation) that is broken.

@scorpion_87 your issue is the same one here: AIDA64 crashes when shutting down

this crash has been misattributed to vanguard but its caused by kerneld.x64 drivers in aida64 builds newer than 5.99.4900, vanguard does not even turn up in the loaded modules (nor is it installed) the vanguard crash occurs the moment you click the sensors page, not on close.

RTSS worked around it by using another cpu time counter, it is not fixed, only microsoft can fix it.

it requires some (3 days minimum) uptime without reboot. kernel memory dump blames MiMappingHasIoTracker trying to read from zero'd memory. I can reproduce it across multiple windows 10 systems, but so far not windows 7. All dump files across multiple user systems that i have seen so far are all sharing vary similar stacks. have seen it on i3 second gen in a toshiba aio touchsystem, and x79 on rampage iv with xeon v2's. rst driver doesn't matter, smart/raid access doesn't matter all that matters is moving from the 4xKB driver in 5.99.4800 to the 67KB driver in 5.99.49xx beta's and later. Here is a list of affected configurations, there is no correlating hardware or driver apart from kerneld.x64 and windows own drivers. Msi Z270 Tomahawk - i5-7600K - Asus GTX 1050TI. MSI B150 Gaming M3 I7 6700 GTX 770 i7-9700K RTX 2080Ti Z390 chipset MB Ryzen 5 3600, MSI B450-A PRO MAX, 16GB DDR4 Vengance 3200 Mhz, 550W PSU, RTX 2070 i3 10105f, Gigabyte B560M DS3H V2, GT 1030 H97m-e ASUS whit an i5 4460 and rx 480 Quosmio DX730. i5 2450, GT 540, 16GB. Single SSD. Almost an untouched install of windows on it, it only runs some dedicated servers. RST 12.9.4.1000 Rampage IV Extreme, 1680v2, GTX 1060, 32GB. RSTe 6.3.0.1022, iaNVME 5.3.0.1005, Samsung nvme 3.3.0.2003 My suspicion is a core windows driver is responsible. Windows 11 does not seem affected? I do have a suspicion that the Microsoft XVDD driver is involved, as the bsod can trigger any point after the event log registers a disk surprise removed event you appears to be confused and mixing some idle bsod issue into this one.

Confirmed, still affects 6.80. oddly, doesn't occur on windows 7. There hasn't been a driver change since atleast 6.33

Same, seeing either that or a hard lock.

disable cpu throttle measurement under Hardware Monitoring.