Summary

  • RAM frequency impacts performance more than latency in most cases.
  • CAS latency is especially useful for speeding up video editing and rendering use cases.
  • Balance frequency and latency based on your needs; 6000MT/s CL30 kits work for most.

Amidst all the talk about the best GPUs for gaming and the best CPUs to pair with them, it's common for us to sideline the intricacies of memory. You might keep things simple by picking one of the best RAM for gaming, but there's more to it than just choosing between DDR4 and DDR5 RAM.

Leaving aside aspects like the memory die used in your RAM, you still need to consider the frequency and latency of the memory kit you're buying. Depending on your use case, either of these factors can matter more than the other.

👁 Image of a Corsair Dominator DDR5 RAM chip
How to easily overclock RAM without BIOS

If you have an AMD Ryzen CPU, overclocking your RAM can be quite simple with Ryzen Master

RAM frequency

The need for speed

This is the spec that most people use to identify a memory kit. While the best DDR5 RAM for gaming can have frequencies of up to 8000MT/s, you can also consider DDR4 RAM for gaming. But, how exactly does frequency affect your RAM's performance?

Well, you can consider frequency or transfer rate as the speed at which data is transferred from your storage to your memory. It's measured in megatransfers per second (MT/s). The faster the rated frequency of your RAM kit, the snappier your system will feel (though at some point you may not feel it). Memory frequency has the most tangible impact on gaming performance and the day-to-day performance of your PC.

The previous-gen DDR4 memory kits ended up with 3000-3600MT/s the sweet spot for most systems, while for the newer DDR5 memory kits, 6000-6400MT/s can be considered the optimum RAM frequency for most applications.

RAM latency

Speed isn't everything

Source: Unsplash

It's true that memory frequency has a bigger impact on performance (in most cases) than latency, but you can't ignore the latter altogether. Memory latency is the time it takes for your RAM to respond to a request. Latency is composed of various factors but we'll limit this discussion to CAS latency, which is denoted by the "CL" or "C" number in your memory specs.

You can check your kit's CAS latency by looking at the Memory tab in CPU-Z. Under Timings, you can find it next to CAS# Latency (CL). This number denotes the number of clock cycles your RAM needs to access a set of data. The actual latency in nanoseconds is calculated by the formula below:

Latency (ns) = (CL * 2000) / Data rate

Data rate is simply the rated speed of your memory kit (3600 or 6000 or 6400MT/s). RAM latency matters more in applications like video editing and rendering. For gaming, latency doesn't have a whole lot of impact on your FPS.

RAM frequency vs. latency

What does the data say?

To illustrate the performance impact of frequency vs. latency, we ran a few CPU benchmarks using CPU-Z, Geekbench, and AIDA64 Extreme. These tests measured the CPU performance as we switched between kits of different frequencies and also between different latencies on the same kit.

CPU-Z

Geekbench

AIDA64 Extreme

DDR5-4800 CL40

  • 1T: 764
  • nT: 10,444
  • 1T: 2,915
  • nT: 16,285
  • Read: 59,066 MB/s
  • Write: 56,779 MB/s
  • Latency: 91.3 ns

DDR5-6400 CL40

  • 1T: 766
  • nT: 9,867
  • 1T: 2,943
  • nT: 17,488
  • Read: 81,473 MB/s
  • Write: 79,972 MB/s
  • Latency: 78.8 ns

DDR5-6400 CL32

  • 1T: 762
  • nT: 9,915
  • 1T: 2,951
  • nT: 17,580
  • Read: 107,230 MB/s
  • Write: 104,389 MB/s
  • Latency: 69.9 ns

You can see the single-thread (1T) and multi-thread scores (nT) for CPU-Z and Geekbench for our 4800MT/s and 6400MT/s kits. The AIDA64 Extreme results show the read/write performance and the associated latency of the kits at different XMP settings.

You can clearly see the impact of memory frequency and latency on the memory test score in AIDA64 Extreme. When going from a 4800MT/s to a 6400MT/s kit at the same latency, there's a 37-40% increase in the read/write score. On the other hand, decreasing the latency from CL40 to CL32 for the same frequency results in a relatively lower 30% increase in the read/write score. This demonstrates the bigger impact that frequency has on memory performance than CAS latency.

The Geekbench CPU test shows a change only in the multi-thread score when moving between kits. And only a frequency increase is able to move the score significantly. The latency improvement doesn't really affect the scores much.

What should you prioritize?

While we've seen RAM frequency affect memory performance much more than latency, you need to strike a balance between the two when buying a memory kit. If you simply want a great kit for gaming alone, a DDR5-6000 CL30 kit is the sweet spot. And if you're looking for a productivity-focused kit, you can consider faster speeds — 6400MT/s to 8000MT/s — based on your budget and platform. The best Intel CPUs are known to support memory kits faster than 6000MT/s more easily than AMD's AM5 CPUs.