Is there any difference between 3200, 3333, and 3600 RAM speeds?

Mykola Lytvynchuk
3 min readFeb 3, 2022

There is a lot of info on the Internet, that RAM speed is crucial for PC performance. So I decided on my own to check this. I already had 2 kits: G.Skill DDR4 16GB 2x8GB 3200MHz Trident Z RGB (F4–3200C16D-16GTZR) and HyperX DDR4 16GB 2x8GB 3333MHz Predator Black (HX433C16PB3K2/16). The next step was to find something faster, so I picked up G.Skill DDR4 16GB 2x8GB 3600MHz Trident Z NEO RGB (F4–3600C16D-16GTZNC).

My setup: AMD Ryzen 5 3600 s-AM4 3.6GHz/32MB BOX CPU and Asus ROG STRIX B450-I Gaming s-AM4 B450 Motherboard.

3200 vs 3333 vs 3600

The first kit was the 3200 MHz kit. We will use it as a standard for all next measurements. I checked it in Cinebench R23, Geekbench 5, and Counter-Strike test.

The next kit was the 3333 MHz kit and the same tests. As we can see there is not much improvement. Also, some synthetic tests are even worse like multi-core performance in Geekbench 5 and Counter-Strike average FPS.

I was worried about the next kit due to not having the best results in the previous test. Let’s check the scores for the 3600 MHz.

Results aren’t impressive at all and in some areas are worse. It led me to tighten up the timings or in other words RAM overclocking.

I started from the safe config and it led to the best scores in Cinebench 5.

I decided to try fast config and it went even better.

Results were much better in synthetic benchmarks but the gaming performance was almost the same. In a conclusion, there is not much difference for gaming to pay for something, which is faster than 3200 MHz. But the synthetic tests for computation work were better at 3600 MHz after tightening up timings, so it could be worth extra bucks.

P.S. I think RAM overclocking for the 3200 kit could achieve similar results as we got for the 3600 kit.

--

--