RAM mixing

ksud

ksud

Disciple
Hi folks.

TL;DR - Does RAM mixing (Brand & Rank) cause any major performance penalty?

I have an opportunity to exchange my current 8GBx2 SODIMM Ram Kit with a Samsung 16GBx1 module. I started using dual monitor setup few months back and 16GB ain't cutting it. I don't have access any offline retailers that can get me an exact module so would have to mix and match RAM. Ideally would prefer to get a Samsung module but can't find any on Amazon that is in stock or under 5k. Crucial seems to be a lottery game. Sometimes you get module with different timings and rank than stated in product description. My laptop has an i7-8750H and I am not sure if this generation is that much affected by rank.

Any one of you have experience using both RAM from same batch and RAM from mixed vendors with 8th Gen? I don't want to leave performance off the table as this laptop would start to show it's age after some years so I'm happy to wait to get a Samsung RAM Module with same Rank and timings.

Attaching below my Kingston RAM kit and my friends Samsungs info.
 

Attachments

  • Kingston.png
    Kingston.png
    25.5 KB · Views: 55
  • Samsung.jpg
    Samsung.jpg
    281.4 KB · Views: 54
As long as the Ram Type and Speed is same, brand does not matter. You can check the current module's speed and type and get another module matching it. Incase you are not getting the same speed and put in a slower speed module, then the system will use both the modules at the slower speed.

MaSh
 
If you aren't buying a 2 pack of matched pair, they may work, depending on your luck, if the MHz and timings are within tolerances. Else they will work at a lower speed and/or higher latency (the lowest in the group).

Sometimes sticks from the same manufacturer but different batches don't play well with each other.
Eg: I was unable to find a matched 4x 16GB sticks for my last build, so I bought 4 single 3600MHz CL18 pieces.
But they would only work at 3200 MHz. Turns out one of them was from a different batch. I replaced that to match the others, so now I have the 4 of them running perfectly at 3600MHz

TL DR: if you don't want to experiment, sell the single stick you have and buy a matched pair. BUT RAM is so fast, any difference in performance is imperceptible outside of benchmarks. More RAM is almost always better than faster RAM if your apps need the extra headroom.
 
I've mixed 4+4G RAM from 2 different brands on my previous laptop and had no issues.
It would just run at the slower rated speed of the two.
 
I've had multiple systems with mic and match RAM, a desktop PC where I had a 256 GB and a 512 GB module, then one where I had a 1 GB and a 2 GB module (those days, I didn't even know about RAM timings), then a laptop where I had 4 GB RAM and added a module of matching speed later (still didn't check timings), my dad's laptop which came with 4 GB and I added 8 GB in the second slot, another laptop of my own which came with 4 GB soldered 2400 MHz and I put 2666 MHz module in it, and finally a Synology NAS which comes with 2 GB soldered and I put 4 GB module in it.

Out of so many systems where I have mixed and matched RAM, the number of times I have had issues because of it is exactly zero.

If, as someone mentioned, the chances of having issues was 50/50, then I'm probably the luckiest guy in the world.

Or it might just be that systems are intelligent enough to account for these things and not create issues. Yes, your RAM might not run in dual channel in some cases, but outside of gaming, that hardly matters at all.
 
Cool. Will wait and see if can get any RAM module with the same speed and timings. Till then I plan to experiment with 8GB + 16GB module.
 
I have been mixing rams in all sort of envi. till date for last 2+ decades and never encountered any issues.
Have mixed corsair/gskill/kingston/transcend/micron/hynix etc.

If the specific mobo doesn't officially support a specific module or a size then nothing could be done.
 
So bought a Hynix ram module with same advertised speed and rank. Was a little let down when it turned out to be 2400mhz rated speed. Did some overclocking and now both my Samsung and Hynix sticks are running at 2666Mhz and CL16 timings. Tired CL15 but I was pushing my luck. CL16 seems stable so far.

Thanks MSI for advanced bios cheat code :kissingsmiling:. Also thanks forum peeps for your inputs:grimacing:.
 

Attachments

  • CURRENT SETTINGS 18dec 2022.jpg
    CURRENT SETTINGS 18dec 2022.jpg
    90 KB · Views: 51
  • CUrrent.jpg
    CUrrent.jpg
    59.7 KB · Views: 41
I have also mixed RAM from different brands and timings with no issues whatsoever. I mixed Corsair vengeance 2*8 GB with a 8 GB Kingston DDR3 (different timings) and then also used 32 GB total. 8*2 Corsair and 8*2 Kingston.
 
Motherboard: msi h110m-pro vd plus. Max ram speed 2133mhz non overclockable.
I have a kingston valueram 8gb 2133mhz. I got a corsair vengence 8gb 2400mhz from my friend and put it in my pc. Main purpose was to increase fps in csgo by utilising dual channel. Well it worked, got 20% average fps increase from 153 to 184 FPS. I did initially witness longer boot time ( i have hdd, no ssd), but then it went back to normal without me doing anything. So in my case there was no negative effect.
I believe there is also a linustechtips video about ram mixing you can check that out too
 
Back
Top