HP V10 RGB DDR4-3600 CL14 2x 8 GB

[Taraquin]

I'm no disagreeing that tRFC has a impact in certain setups. What I am saying is that it cannot be said as a blanketed statement as you are implying. I applies more to 1920x1080 resolution with lower graphical settings, but some games like Metro: Exodus doesn't care at all what the tRFC value is.

Generally it has impact in most games if cpu bound, but of course there are exceptions. Lower rfc means lower latency, in aida 10 tics lower rfc is about 0.1ns. Most, but not all games scales good with latency if cpu bound that is.

My original inquiry was that if you tune B-die you should lower rfc a lot vs xmp since all B-die can handle low rfc (140-160ns@1.4V depending on bin) and that is the timing that really separates B-die from all other dies.

 

[the ram]

Rot dot? xD

 

[ir_cow]

Rot dot? xD

shhh, no one noticed

 

[Mussels]

We will just have to agree to disagree on this one. Your mileage may vary.

For example, If all b-die kits could do the same than that Ryzen Calculator would be perfect, yet people are taking those values as absolute. Its a great tool, but is more of a guideline. Same goes with RFC. It is also not of significance outside of benchmarks to go from 560 to 350. The other timings like tFAW is more important for tuning.

I'm just posting to backup up Taraquin as actually knowing a lot about DDR4 overclocking, talked to him in several threads and gotten a lot of advice from him
(No, he's not magic or always correct - but his claims are definitely backed up by personal experience)

 

[ir_cow]

I'm just posting to backup up Taraquin as actually knowing a lot about DDR4 overclocking, talked to him in several threads and gotten a lot of advice from him
(No, he's not magic or always correct - but his claims are definitely backed up by personal experience)

I'm not trying trying to discredit @Taraquin . I agree with him that B-Die can do more. However my counter argument is two fold. First no overclock is guaranteed and that these lower values cannot be blindly applied without some sort of stability testing to confirm it is good for YOUR personal setup. Secondly, while this fine-tuning of the sub-timings can be beneficial, I think it should be stated in context. In short, besides synthetic benchmarks, I do not believe that every program/game will reap the rewards. I am not personally going to fight the MB tooth and nail over 1 FPS.

This wraps around to why I choose to leave the sub-timings a little looser. This is because most readers that overclock memory will just input these values and never check or know how to check for 24/7 stability. I do not suggest it and I am guilty of this when I was less experienced. People will do this regardless of my suggestions.

Once again no overclock is guaranteed, but if you are attempting it, it is better to start a bit higher and work your way down in timings. I rather provide an average real-world timings, instead of ones that require a fan to keep the B-Die below 45c. I don't see the point of going all-in on a review when really tightening down the timings isn't directly usable for most people. This is why the Ryzen Calculator is a great guideline, but more often than not, even the "safe" values leave the computer in a non-bootable state.

 

[megaclite]

Why not buy the 14-14 set at this price though.

$120

OLOy Blade 16GB (2 x 8GB) 288-Pin DDR4 SDRAM DDR4 3600 (PC4 28800) Desktop Memory Model ND4U0836144BRQDE - Newegg.com

G.SKILL Ripjaws V Series 16GB (2 x 8GB) 288-Pin DDR4 SDRAM DDR4 3600 (PC4 28800) Intel XMP 2.0 Desktop Memory Model F4-3600C14D-16GVKA - Newegg.com

DDR4 is so cheap now, even the best 3600c14-14 stuff

Oloy 3600 14-14 1.40 highest quallity B-die, better than 90% of g.skill stuff
Mine can do 4000c14 1.55v

 

[Mussels]

I'm not trying trying to discredit @Taraquin . I agree with him that B-Die can do more. However my counter argument is two fold. First no overclock is guaranteed and that these lower values cannot be blindly applied without some sort of stability testing to confirm it is good for YOUR personal setup. Secondly, while this fine-tuning of the sub-timings can be beneficial, I think it should be stated in context. In short, besides synthetic benchmarks, I do not believe that every program/game will reap the rewards. I am not personally going to fight the MB tooth and nail over 1 FPS.

This wraps around to why I choose to leave the sub-timings a little looser. This is because most readers that overclock memory will just input these values and never check or know how to check for 24/7 stability. I do not suggest it and I am guilty of this when I was less experienced. People will do this regardless of my suggestions.

Once again no overclock is guaranteed, but if you are attempting it, it is better to start a bit higher and work your way down in timings. I rather provide an average real-world timings, instead of ones that require a fan to keep the B-Die below 45c. I don't see the point of going all-in on a review when really tightening down the timings isn't directly usable for most people. This is why the Ryzen Calculator is a great guideline, but more often than not, even the "safe" values leave the computer in a non-bootable state.

That's actually what i ran into, having cheaper hynix ram (2x32GB didn't help) his estimates for timings didn't work for me initially.

I agree with going slightly looser for the reviews sake, in case users copy... a memory review isn't an overclocking how-to with it's millions of intricacies

 

[Emboldi]

Is anything justifying spending money on 3600 cl 14 coming from 3200 cl16 at this point? (dont want to switch the 5900x)

 

[ir_cow]

Is anything justifying spending money on 3600 cl 14 coming from 3200 cl16 at this point? (dont want to switch the 5900x)

Depends on your personal opinion. CL14 does have a advantages if you are frame chasing with the right hardware configuration.

 

[Mussels]

Is anything justifying spending money on 3600 cl 14 coming from 3200 cl16 at this point? (dont want to switch the 5900x)

differences in anything other than benchmarks would be very very small
without info on the rest of your specs or even the sizes you're after, we cant say much


four sticks of 3200 C16 would perform equal or better than two sticks of 3600 C14 for gaming, if they were all single rank