Similar questions have been asked many times, I know. I've researched forums all across google and haven't found exactly what I'm looking to know so I'll ask here.
latency vs speed
It's my understanding that the last digit in the timing is essentially the whole number of clock cycles each call takes.
Specifically what I'm looking at is
8-8-8-26 @ 1800MHz
9-9-9-28 @ 1866MHz
I'm trying to do the math and figure out which will give better results based on the numbers. Here's how I'm working through it:
Each cycle of 1800MHz takes (1/1800)seconds. 26 clock cycles at this speed results with a time of 26 * (1/1800) = (26/1800) = 0.0144 s
Each cycle of 1866MHz takes (1/1866)seconds. 28 clock cycles at this speed results with a time of 28 * (1/1866) = (28/1866) = 0.0150 s
For my example it would seem that the slower RAM would give the better results. Is this correct?
Also - if the controller is pulling a large chunk of data, does this same lag occur at each bit? just each chunk of data? I know 'burst' comes into play here, but I don't understand exactly how.
If the chip is trying to pull a 10bit chunk from RAM and it incurs this latency at each bit, the latency would seem to be the more important factor. If it only incurs the penalty on the first bit and the subsequent addresses are read with only one cycle latency (which I think is what 'burst' is?) then speed would be more important.
I know the difference here would me extremely minimal, but I'd like to really understand what's going on before choosing.
Any help would be greatly appreciated, even if it's just pointing me to a resource that can guide me.