• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Combining different ram - will it work ?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

SentinelAeon

Registered
Joined
Apr 2, 2020
Location
Europe
So i am trying to combine a ram into a dual channel:

Channel 1: - 4GB 1600Mhz 9-9-9-24 brand A
- 2GB 1600Mhz 11-11-11-28 brand B

Channel 2: - 4GB 1600Mhz 9-9-9-24 brand A
- 2GB 1600Mhz 11-11-11-28 brand C

The 4GB sticks are identical, but the 2GB ones aren't. Obviously, they will all work with 11-11-11-28 timings. But my question is this: If i put memory in this configuration and run all mem tests i know of, and they all pass. Will i then be able to be reasonably sure that all will work well ? Or is there a big chance that even though all tests pass, that because the 2GB sticks aren't identical brand, there will be problems ?
 
If you make it work and pass tests then all will be fine. Usually it's more about memory design than brands.
It's recommended to use lower capacity modules in first slots but it's also not a rule as it can work either way. Usually motherboards use SPD/XMP settings from first module(s) and set it like that in main and sub-timings for all others. In theory if there is any compatibility issue then use the slowest RAM in the first slot so it uses these "invisible" settings in all modules.
When you mix various capacity and IC modules then you have to check it on your own as there is never clear answer. However, when memory is higher loaded for longer and doesn't show any errors then all is fine. You can check something like 7-Zip benchmark to fill the RAM in ~90% and perform more complicated operations. AIDA64 stability test or anything similar is fine too.
 
Thank you for this tip, i never thought it works like this. So i will put into first slot the weaker ram with higher timings, if it doesn't work i will turn it around. Lets hope it will be fine. This ram is 1.5V default, is 1.6V ok for 24/7 in case with good fans ? I plan to tempt my faith further and either lower all timings from 11-11-11-28 to 9-9-9-24 and/or go for that 2133Mhz.
 
DDR3 works fine up to at least 1.65-1.70V. I had some kits that didn't like more than 1.75V but nothing happened at a higher voltage. Above some voltage they were throwing bluescreens during windows load. Once lowered then all was again fine. But that's above 1.75V on a specific, not even popular IC. Safe 24/7 voltage is at least 1.65V on all available DDR3. It shouldn't need any additional airflow at this voltage.
 
Thank you. I will test the ram soon and i will limit it to 1.6 or 1.65V. As i said, the weaker sticks are 1600Mhz 11-11-11-28 and the better 2 are 9-9-9-24. So what would be your opinion on which path i should try to accomplish:

1) try to get all 4 sticks to be 1600Mhz 9-9-9-24
2) try to get all 4 sticks to be 2133Mhz 11-11-11-28 (or even more loose, as long as it works on 2133Mhz)

I dont need performance in games right now, i like windows snappy and i encode H.265 video with my cpu a lot
 
I have some news and some questions. I tried all the sticks i have, had some disapointments and some surprises (disapointment: 1600 9-9-9-24 balistic heatsinked ram vs 1600 11-11-11-28 hyundai ram that came with i5 3xxx office pc ... both at 1.65V 13-13-13-32 2133Mhz, balistic didnt boot, hyundai did, ran memtest for a while then threw out errors). Also susprise, 1333Mhz 9-9-9-24 low profile ram, managed to get it to 1866 11-11-11-28 at 1.65V.

I have 2 questions, i hope u can help me with them:

1) Should i choose 1600Mhz 9-9-9-24 or 1866Mhz 10-10-10-26. This are the only options i managed to get (meaning i cant get it to run at 1866Mhz with lower timings)
2) Does it make sense to try to lower individual timings, for instance 10-9-9-25 vs 10-10-10-26 ? Or should they be the same ?
 
No one can answer that question. It's all trial and error, dude.
 
Back