Thanks. Please try this build and let me know the result: www.hwinfo.com/beta/hwi64_765_5256.zip
A new Debug File of suspend/resume might be needed from this build.
A new Debug File of suspend/resume might be needed from this build.
High CPU load
- Thread starter waleedov
- Start date
That was what I could do in this case. The rest is due to how various sensors are designed and how much resources it takes to read them. To reduce the the load further you will need to disable some sensors.
I also don't think that the Power usage in Task Manager is a reliable indicator.
I also don't think that the Power usage in Task Manager is a reliable indicator.
Thanks! I can confirm that your build 5256 appears to solve all of the observed issues for my 12-th gen Alder Lake CPU on my Gigabyte Z690 mobo running Windows 11 Pro.
I did end up disabling the "DDR5 DIMM..." monitoring as, while not as heavyweight as the Renesas stuff, still stood out as being resource-intensive to watch - I don't suppose you can address that too?
I did end up disabling the "DDR5 DIMM..." monitoring as, while not as heavyweight as the Renesas stuff, still stood out as being resource-intensive to watch - I don't suppose you can address that too?
Hi Martin, I just noticed that HiNFO64 is back to sucking up more than an entire core's worth of CPU on my same system (Intel 12-th gen Alder Lake) - and the solution from last time, disabling the DDR5 sensors, is still in place.
This is the automatic update to version/build v7.73-5375 - any ideas what I may have to disable now, or possibly even better solutions?
EDIT: I just noticed which ones are causing the problem this time - it is two sensor groups, both showing the same name: "GIGABYTE Z690 AORUS MASTER (Renesas RAA229131)". If I disable monitoring both of these two groups, then the HWiNFO64 cpu usage drops back to the "hardly there" level.
This is the automatic update to version/build v7.73-5375 - any ideas what I may have to disable now, or possibly even better solutions?
EDIT: I just noticed which ones are causing the problem this time - it is two sensor groups, both showing the same name: "GIGABYTE Z690 AORUS MASTER (Renesas RAA229131)". If I disable monitoring both of these two groups, then the HWiNFO64 cpu usage drops back to the "hardly there" level.
Last edited:
In case you're running some other monitoring/tweaking tools (including GIGABYTE SIV), try to close them.
If that won't help I'll need to see the HWiNFO Debug File with those sensors enabled to see if there's some issue.
Nope, only GPU-Z... funny thing about these sensors: are they new in HWiNFO64? They seem to be all about power delivery to the CPU. Also, they don't seem to appear when my app enumerates all your sensors in the shared memory segment(?).
And if you really want the debug file, please point me to the instructions for enabling it.
No, these sensors were added a couple of versions back.
www.hwinfo.com
IMPORTANT - Read this before submitting a report
In order to submit a bug report, please include the following information: HWiNFO32/64 Report File (HTM or TXT preferred) HWiNFO32/64 Debug File (HWiNFO32.DBG or HWiNFO64.DBG) In case the application crashes or you are not able to create a Report File, please submit only the Debug File. This...
www.hwinfo.com
Thanks for the link, maybe I won't lose it.
But you may not need your "Debug File"... if the end result of this query is the same as last time - "reading the DDR5 info is expensive, disable it if you are concerned about the CPU usage" - then I can already see what to do here: disable the two "GIGABYTE Z690 AORUS MASTER (Renesas RAA229131)".
Basically, I only have 5 "expensive" sensors in my setup: the 2 DDR5 @ 22ms EACH, the 2 "Renesas" @ 11ms EACH, and the "GPU [#0]: NVIDIA GeForce RTX 4070: MSI RTX 4070 Gaming X Trio" @ 18ms - 28ms (variable, usually down around 18-20ms).
Here is something interesting - I *just* let Windows install the latest "Insiders Preview Cumulative Update" - and now, as I re-enable these "expensive" sensors, the CPU energy consumption only goes up incrementally - NOT suddenly demanding a whole core's worth of power! Maybe they cleaned up something in the VM emulation layers, which I think Win 11 is using all of the time... maybe a contention issue is gone now? This newest build is "22635.3276", BTW.
In any case, I can still disable these sensors if I choose to, or not - but I am not driven to disable them, if you see the difference.
But you may not need your "Debug File"... if the end result of this query is the same as last time - "reading the DDR5 info is expensive, disable it if you are concerned about the CPU usage" - then I can already see what to do here: disable the two "GIGABYTE Z690 AORUS MASTER (Renesas RAA229131)".
Basically, I only have 5 "expensive" sensors in my setup: the 2 DDR5 @ 22ms EACH, the 2 "Renesas" @ 11ms EACH, and the "GPU [#0]: NVIDIA GeForce RTX 4070: MSI RTX 4070 Gaming X Trio" @ 18ms - 28ms (variable, usually down around 18-20ms).
Here is something interesting - I *just* let Windows install the latest "Insiders Preview Cumulative Update" - and now, as I re-enable these "expensive" sensors, the CPU energy consumption only goes up incrementally - NOT suddenly demanding a whole core's worth of power! Maybe they cleaned up something in the VM emulation layers, which I think Win 11 is using all of the time... maybe a contention issue is gone now? This newest build is "22635.3276", BTW.
In any case, I can still disable these sensors if I choose to, or not - but I am not driven to disable them, if you see the difference.