Adding 2 new Ram Sticks, stability

[A Computer Guy]

Thank you, i will study this.
Going from CL 18 to CL 16 is also overclocking?

I think so.

I was reading and seems that i would benefit more from latency reduction that Mhz increase, cause of infinity fabric 3600mhz is the sweets spot?

It depends. On one end of the scale is latency on the other is bandwidth. You want to find a happy place that is an acceptable tradeoff between the two for your needs.
If you can scale your memory kit up in bandwidth without increasing (or minimizing the increase) the latency in nanoseconds then that is a win, win.

For my Corsair DDR4-3200 memory kit on my Ryzen 3800x OC to DDR4-3800 is the sweetspot in terms of overall bandwidth since I can run a 1900Mhz IF (infinity fabric) however DDR4-3733 (with matching IF) end up being a better match with less tradeoffs in increasing latency (in nanoseconds). Before I discovered Dram Calculator I used to recalculate the timings in nanoseconds to try and find settings that might work before trying to tighten them. I hear IF frequency is better in Zen3 CPU's.

About the calculator, does it work properly for ryzen 5000 ? I read other guides and they say only goes up to 3000 the "auto-calculators"

It should be good enough however inputting the correct memory type to get good recommendations can be an issue if you don't know, or if the SPD info in the ram is incorrect when reported by thaiphoon.

My Corsair kit CMW32GX4M2C3200C16 (ver 4.32) (that reports they same part number K4A8G085WB-BCPB as yours) . In DRAM Calculator "Samsung OEM" used to be the closest match in terms of compatible settings. This is why I was asking if you know the version number of your kit since you might have a similar experience. Later from an OC forum I found kits of this version seem to behave more like Samsung c-die not b-die as reported by thaiphoon. I had to actually lower the DRAM voltage to 1.33v to get the higher overclocks.

I know it's a crap benchmark to use but I use User Benchmark as a quick and dirty test for ram overclock to see how it is doing, in particular to see the tradeoffs between the ram and cpu benchmarks. If the ram OC hurts CPU utilization then I rebalance the ram OC to get better CPU utilization.

 

[milewski1015]

Going from CL 18 to CL 16 is also overclocking?

Yes. Even though you're not increasing the memory frequency, you're tightening the timings to reduce latency. I interpret overclocking in this day and age to be any instance of pushing something beyond it's specifications to increase performance.

 

[repije]

I think so.

It depends. On one end of the scale is latency on the other is bandwidth. You want to find a happy place that is an acceptable tradeoff between the two for your needs.
If you can scale your memory kit up in bandwidth without increasing (or minimizing the increase) the latency in nanoseconds then that is a win, win.

For my Corsair DDR4-3200 memory kit on my Ryzen 3800x OC to DDR4-3800 is the sweetspot in terms of overall bandwidth since I can run a 1900Mhz IF (infinity fabric) however DDR4-3733 (with matching IF) end up being a better match with less tradeoffs in increasing latency (in nanoseconds). Before I discovered Dram Calculator I used to recalculate the timings in nanoseconds to try and find settings that might work before trying to tighten them. I hear IF frequency is better in Zen3 CPU's.



It should be good enough however inputting the correct memory type to get good recommendations can be an issue if you don't know, or if the SPD info in the ram is incorrect when reported by thaiphoon.

My Corsair kit CMW32GX4M2C3200C16 (ver 4.32) (that reports they same part number K4A8G085WB-BCPB as yours) . In DRAM Calculator "Samsung OEM" used to be the closest match in terms of compatible settings. This is why I was asking if you know the version number of your kit since you might have a similar experience. Later from an OC forum I found kits of this version seem to behave more like Samsung c-die not b-die as reported by thaiphoon. I had to actually lower the DRAM voltage to 1.33v to get the higher overclocks.

I know it's a crap benchmark to use but I use User Benchmark as a quick and dirty test for ram overclock to see how it is doing, in particular to see the tradeoffs between the ram and cpu benchmarks. If the ram OC hurts CPU utilization then I rebalance the ram OC to get better CPU utilization.

My memory is 4*8GB of Corsair 3600Mhz CL18, ver 4.32 on the stick althouht i bouth 2 packs separetly so the part number is a bit different from yours.

Need to test alot of stuff I see, so many boots waiting for me xD

Yes. Even though you're not increasing the memory frequency, you're tightening the timings to reduce latency. I interpret overclocking in this day and age to be any instance of pushing something beyond it's specifications to increase performance.

got it thanks