Two of these RAM modules are not like the other

So recently I decided to get another 16GB of RAM for my primary rig. To ensure minimal performance degradation (as when mixing kits often memory will revert to the lowest common denominator (not to mention matching modules looks nice). I got the exact same kit referring to the model number on my existing receipt to ensure a match. The kits in question are the G.Skill Sniper X 16GB DDR4 3600Mhz kits (F4-3600C19D-16GSXWB), both kits share this model number and all 4 modules have the part number F4-3600C19-8GSXWB so there’s little doubt I got sent the wrong kit. The thing is they’re not the same. One pair has Samsung chips, the other SK Hynix. Between the two sets of modules, SPD data also varies. In my instance it’s not terrible, the XMP profile works perfectly and timings remained the same for that singular profile, I do however have some concerns for those who operate at JEDEC profiles as they differ in frequency and timings/sub-timings.

Now buying the exact same memory model now doesn’t always mean you’ll get the exact same memory, but that doesn’t always have any meaningful impact which is thankfully the case in my instance. I do however have some concerns regarding QVL and setups which are more finicky with memory, although I wonder how valid they are given that at XMP the frequency, timings, and sub timings are all identical; I suspect very few people would be running these on their JEDEC SPD profiles, also I’m fairly certain that the timings are tighter/frequencies higher on the new modules and as such, they should generally interoperate fine.

Below are a few screenshots of all the timings followed by a rough benchmark, I’ve included both CPU-Z and AIDA64 as I find the CPU-Z timings easier to read however unfortunately it only displays 4 of the 8 profiles on the modules.

So long story short I bought the same model memory and my modules were different! Just thought some people might find this to be an interesting quirk that’s, unfortunately, happening more and more often with various PC hardware. Thankfully this one isn’t a bait and switch, as has occurred with some SSD’s as the new modules perform as specified and likely would exceed the performance of the original modules if used alone.

Module pair 1 (SK-Hynix):

Module pair 2 (Samsung):

A rough benchmark of them running happily at XMP (rough as I didn’t clean boot or control the environment in any meaningful way beyond closing running applications).