I'm using a Ryzen 5 5600X with 16 GB of DDR4-3200 memory in a 2x8 GB kit. I was given another 2x8 GB DDR4 kit rated for 3600 MHz, and I'd like to use all 32 GB because some games are maxing out my current memory and causing inconsistent 1% lows. Since the kits have different rated speeds, would it be better to overclock the 3200 MHz kit to 3600, or run the 3600 MHz kit at 3200? I've overclocked a CPU before but haven't tuned RAM, so I'm also wondering whether 3600 MHz is meaningfully better on a Ryzen 5 5600X or mostly a small benchmark difference.
3 Answers
The memory chips and rank layout can affect how well the kits work together, especially if the modules use different ICs or a mix of single-rank and dual-rank designs. Identify the exact models and rated timings if possible, then test with a memory stress test after each change. Avoid blindly increasing voltage, since some memory ICs can be damaged by settings that are safe for other kits.
Start by running both kits at 3200 MHz. A mixed-memory setup is more likely to be stable when the faster kit is slowed down than when the weaker kit is pushed beyond its rating. Check the rated timings for both kits and use settings the slower one can handle. If it works reliably, you can experiment with 3466 or 3600 later, but stability matters more than the small performance gain.
On a Ryzen 5 5600X, DDR4-3600 can be a nice target because it often pairs well with the CPU’s fabric clock, but the real-world improvement over DDR4-3200 is usually modest. With mixed kits, 32 GB running consistently at 3200 is generally preferable to chasing 3600 and getting crashes, errors, or worse latency. Matching timings and maintaining a 1:1 fabric ratio can matter as much as the headline memory speed.
That makes sense. I’ll compare the timings first and try the combined 32 GB at 3200 before attempting anything faster.