RAM timings affect how quickly your system accesses data, influencing stability and performance under load.
RAM timings affect how quickly your system accesses data, influencing stability and performance under load.
I’m weighing two options: one with a 3600Mhz clock and another with a 3200Mhz clock. I previously had a 3600Mhz unit but wanted more RAM, now considering upgrading to double it. The specs show the 3200Mhz version runs about 10 nanoseconds per operation versus roughly 8.89 for the 3600Mhz model. Subtimings might have a minor impact, but since the 3600Mhz is similar in speed, the difference seems small.
It's unclear how this measurement applies here. A faster device doesn't automatically mean it matches the claimed speed.
Other things are important, though they aren't as significant as the main point.
I recognize your point, yet there appears to be a lack of clarity, and I haven’t located any relevant information online.
There are many tests being conducted that seem irrelevant, like checking how an RX 5700 XT handles a specific PCI-E gen.4 MOBO. It doesn’t seem to be the main issue. Probably no one is really paying attention, even those who push their RAM to its limits. This is surprising because there’s a clear difference in performance that matters.
Performance depends on both raw speed and timing. Certain tasks require precise timing while others prioritize speed. Generally, stricter timing tends to reduce speed. Timings are usually what people mean by subtimings, as they are the main focus. CAS, tRCD, tRP, and tRAS are key metrics: https://www.gamersnexus.net/guides/3333-...d-trp-tras
These aren't minor adjustments; they're main settings. The smaller ones are just supporting. Regarding your query, the 3600MHz model is the superior choice.
Generally, those settings have minimal effect on performance. They’re considered baseline values rather than dynamic factors. Even though they’re used in latency calculations, changing them beyond the core numbers doesn’t alter the actual measured delay.