Horrific performance boost upgrading RAM from 16 to 24

Hello.
A few months ago unfortunately I decided to upgrade to Windows 11 from Windows 10.
Acer Notebook brand new, Intel generation 11th, 16 GB RAM, Samsung 980 Pro as NVME SSD. Almost the best affordable for business use (and even more).
At once performances were halved, or worse. Really horrible sensation while working... pure rubbish, mostly dealing with files, but not only. I mostly work with cloud ERP for accountancy, Office 365, shared files on OneDrive and plenty of browser tabs (Chrome and Firefox). Lots of windows anytime.
Yes, I know about the Windows 11 problem with NVME SSD.
On Task Manager the total RAM occupied during work was about 8 max 10 GB, on 16 total.
Two or 3 days ago I decided to upgrade the RAM... just because. It wasn't very expensive, so I bought a 16GB board and today I substituted the free 8GB RAM board with the new one, bringing total RAM to 24 GB (first 8 GB are soldered on the motherboard), and... boom! Now it seems that the OS has come again to life! Now the responsivity is almost again the one I had with Windows 10... the difference is really HUGE.
The old RAM board is Samsung, the new one is Crucial, but both are DDR4-3200.
Now with some browser tabs, Outlook, a couple a file explorer opened the RAM usage has risen up to 10-11 GB, but it's completely another way to work, guys!
Someone of you is able to explain this strange behaviour? Anyway it seems I had enough room left in total RAM even before the upgrade and... everyone says that Windows 11 should be OK with 16 GB of RAM. I'm really struggling to understand.

Thanks in advance and kind regards
Alex

Do you actually need all of the services and backgroun applications that are eating up your memory? With a few minor adjustments in both Windows 10 and 11 (I didn't notice any increase in RAM usage), my machines don't use that much. I have 32GB in the machine I'm typing this on, and with Firefox open (with 10 tabs) and task manager, I'm only using 4GB (none compressed). I noticed similar performance on a 5 - 6-year old laptop I gave my wife. It has only 8GB but with normal usage I rarely saw it even hit 50%. Did you clean all of the Acer and other bloatware off of it? Good to see the added RAM worked though.

@AlexDeLarge68


I'm thinking the most likely explanation is that the "not soldered" 8GB stick was dead/dying.
It's probably just coincidence that it happened when you were moving from Win 10 to Win 11.
And when you added the 16GB stick, you gave the computer enough RAM to work with.

abactuon said:

Beware of the single rank x16 RAM kits sneaked in some laptop models! The 1Rx16 modules have significantly lower bandwidth compared to the 1Rx8 variants

The more bank groups a memory module has, the higher the bandwidth is. Contrary to what the naming convention may suggest, the single rank x8 modules have twice as many bank groups compared to the single rank x16 variants. Laptop OEMs unfortunately never reveal the ranking information.

www.notebookcheck.net

Click to expand...

Ehm... I thought I made myself clear but it seems I mistook.
It happened exactly the contrary:
- 8 GB RAM soldered on motherboard
- 8 GB RAM DDR4-3200 brand Samsung installed from purchasing.
- 16 GB RAM DDR4-3200 brand Crucial installed in place of Samsung 8 GB.
- Perfomances more or less taken off after RAM upgrading.
What you posted would lead to an opposite conclusion.

AlexDeLarge68 said:

Ehm... I thought I made myself clear but it seems I mistook. :)
It happened exactly the contrary:
- 8 GB RAM soldered on motherboard
- 8 GB RAM DDR4-3200 brand Samsung installed from purchasing.
- 16 GB RAM DDR4-3200 brand Crucial installed in place of Samsung 8 GB.
- Perfomances more or less taken off after RAM upgrading.
What you posted would lead to an opposite conclusion.

Click to expand...

You are just confusing "x8" with "8GB" and "x16" with "16GB".

AlexDeLarge68 said:

Now with some browser tabs, Outlook, a couple a file explorer opened the RAM usage has risen up to 10-11 GB,

Click to expand...

This is yet another thing that tends to confuse a lot of people. The SysMain (former SuperFetch) service causes the total memory usage to be not in any way representative of how much memory is allocated to other processes running on the system. Unless you actually are using the kind of programs that can benefit from having more physical RAM to work with, the difference in performance either will be minimal to none or may in fact turn out to be a deterioration instead of an improvement. This greatly depends on things like the actual speed and timings of the physical RAM modules in question and on how well the characteristics of the two different modules (or of the module and the bank that is soldered to the motherboard) are matched to each other as well as are matched to other performance related factors inherent of the CPU/chipset and various software optimizations, etc..

Not too long ago I replaced my laptop's stock 2×4GB SK Hynix 3200MHz C22 DDR4 SODIMM Dual Channel Kit with a 2×8GB G.Skill Ripjaws F4 kit of the same speed and timings. Both kits are Single Rank. There is no performance gain that can be noticed, aside from the one that is resulting from the fact I can now store a bigger amount of data on my Romex Primo Ramdisk, as none of my programs are using more RAM. I mean, I know that there exist several many programs that do─it's just that I don't use them.