GPT4All v3.4.2: CPU usage up and down 15% on an i7 depending on GPT4All running in fore/back/ground

[SINAPSA-IC]

I don't know if this is a bug or a feature, in any case it is weird and unexpected, so it (d)evolves into an Issue; to wit:

There are 3 LocalDocs Collections (of PDF files only) being Embedded on CPU (explicitly set in the combobox) at the same time*

• ofc**
  ** still

• While GPT4All is running in BACKground/another program has the focus/or GPT4All is minimized:
  Memory is at 34% and CPU at 29%, with the fans at cruising (regular) speed

• While GPT4All is running in FOREground/maximized/visible/has the focus:
  Memory is at 34% and CPU at 45%, with the fans noticeably/audibly running full speed

• It does not matter which page/view is active (LocalDocs, Chats etc.) - the behaviour is the same.

• the Embedding goes on while in the background as well as while in foreground; however, a 16% jump in CPU cannot be attributed to redrawing/updating the interface (the ProgressBar and the text x%), or can it?

Steps to Reproduce

(to reproduce on my machine, that is)

• 3 LocalDocs Collections are being Embedded
• keep GPT4All in foreground: fans can be heard loudly
• keep GPT4All in background: fans' sound/speed subside

Expected Behavior

Consistent behaviour of the Embedding procedure; as for now, from this^ it looks like there are inconsistencies generated by the foreground status of the program or by the contents of current page/view.

Your Environment

• GPT4All version (if applicable): v3.4.2
• Operating System: Windows 11, i7-12650, Intel GPU
• Chat model used (if applicable):

It may be only happening on my machine with this i7 CPU.

Also, on an i5, the CPU during Embedding is constantly at 6x%-7x% regardless the state of program's window.

When changing the settings of Performance for Programs or Background apps, no sizable difference occurs. So it may not be the Prioritize Foreground setting only, in Windows.