Id take 3200 C14 over 4400 CC19 any day of the week. If you can get that consider yourself fortunate.
Your problem is not the memory or the system but that fact that you bought mixed kits, Huge no no. Doesn't matter if they are the same part number or not, there are probably a thousand or more posts in this forum alone on the same issue. If you an remember which belongs to what kit use just those and Ill bet you get better results. If you need 4 sticks, buy a 4 stick kit.
Your problem is not the memory or the system but that fact that you bought mixed kits, Huge no no. Doesn't matter if they are the same part number or not, there are probably a thousand or more posts in this forum alone on the same issue. If you an remember which belongs to what kit use just those and Ill bet you get better results. If you need 4 sticks, buy a 4 stick kit.
Works the same everywhere. Unmatched kits are a crap shoot at best. Often they wont even post.
I wouldn’t sweat it at 1.4 volts or even 1.5 if you have good airflow or active cooling.*
The faster the memory gets the more succeptive to idiosyncratic behavior it has become. Back when 1600 MHz was the king of the hill it didn’t matter much so long as you bought sticks with the same timings. *The process is different now. After the sticks are made they get binned by their characteristics and speed capabilities. Then they get binned further to exact frequency and timings. Then they get matched up and tested together and sold as kits. You think you have it bad, look at the offerings for 8x8GB kits compatible with X299! *
To make matters worse it seems it’s more often than not memory won’t run at XMP even with a good matched kit and have to be manually tweaked. Sometimes if you simply change your CPU clock a little they will no longer work or they will start working at XMP. Like I said the faster things get the more finicky they become. Just remember the Hz is only half of the equation. 3200 at CL 14*is actually slightly *faster than 3333 *CL15.*
Equation for time to charge
CL/HZx1000= time to charge in milliseconds
14/3200x1000=4.375 milliseconds
15/3333x1000=4.500 milliseconds
Like I said I’d be all over 3200 a C14!!*
The faster the memory gets the more succeptive to idiosyncratic behavior it has become. Back when 1600 MHz was the king of the hill it didn’t matter much so long as you bought sticks with the same timings. *The process is different now. After the sticks are made they get binned by their characteristics and speed capabilities. Then they get binned further to exact frequency and timings. Then they get matched up and tested together and sold as kits. You think you have it bad, look at the offerings for 8x8GB kits compatible with X299! *
To make matters worse it seems it’s more often than not memory won’t run at XMP even with a good matched kit and have to be manually tweaked. Sometimes if you simply change your CPU clock a little they will no longer work or they will start working at XMP. Like I said the faster things get the more finicky they become. Just remember the Hz is only half of the equation. 3200 at CL 14*is actually slightly *faster than 3333 *CL15.*
Equation for time to charge
CL/HZx1000= time to charge in milliseconds
14/3200x1000=4.375 milliseconds
15/3333x1000=4.500 milliseconds
Like I said I’d be all over 3200 a C14!!*
