What is your max infinity fabric on ryzen 5000? September 2021 edition

[Ibizadr]

1.45V is good for my suggestions


Error 0 is usually too low voltage on ram. Try upping by 0.02V, usually fixes it. 15 can be several things, tRFC maybe, but that can often be improved by a bit more volt.


Is the ram in the A2 and B2-slot? Usually this is a problem uf A1-B1 is used as they are the worst slots.

No. they are on a2 and b2. It's weird need to test it without xmp to see if I got errors.

 

[schidddy]

@Valantar mostly what i see is that u defintly get better more efficient results if u tend to clock your cpu at lower speads like 4.5 or 4.6ghz allcore. I am defintly lucky with 4.6ghz at 1.16v under full load, but i see that there is one core not willing to clock higher with slight increased voltage. 4.7ghz is not stable even at 1.25v, maybe needs as high as 1.3v and then the efficency is already getting lost and i get near the power consumption like using pbo. i think pbo always uses high voltages on all cores to maintain high boost posibilty on as much as cores as possible if the best cores are beeing used by some tasks and as i said to be able to get highest boosts for every user as the silicones are never the same

 

[Taraquin]

@Valantar mostly what i see is that u defintly get better more efficient results if u tend to clock your cpu at lower speads like 4.5 or 4.6ghz allcore. I am defintly lucky with 4.6ghz at 1.16v under full load, but i see that there is one core not willing to clock higher with slight increased voltage. 4.7ghz is not stable even at 1.25v, maybe needs as high as 1.3v and then the efficency is already getting lost and i get near the power consumption like using pbo. i think pbo always uses high voltages on all cores to maintain high boost posibilty on as much as cores as possible if the best cores are beeing used by some tasks and as i said to be able to get highest boosts for every user as the silicones are never the same

Still if you tune your curve optimizer max it will generally beat allcore OC.

Here is my scores:
4.4@1.05V 60-65W load
4.5@1.09V 65-70W load
4.6@1.15V 75-80W load
4.7@1.23V 90-100W load
4.8@1.32V 115-125W+ load

Curve optimizer +200 pbo and -30x4/-29x2
4.85@1.34V single core
4.57-4.63@1.16-1.17V@76W

In idle temps with PBO+CO is 40-45C and voltage 0.83-0.85V.

The most comparable 4.6GHz is 45-50W idle and 1.15V.

As you see CO+PBO is superior to allcore OC since you get better single core at lower temp/consumption and very low temp/consumption in idle.

 

[schidddy]

Trc 48, tRFC 264, 272 or 280, twr 16, trtp 8, wtrs 4, wtrl 8 or 10, scl 4 on both. Try that, it should improve your scores quite a bit.

slightly better read speeds and more or less same latency, asking myself if i should use this setting as my daily or go for cl15 or even 14.. but as rares495 wrote i need at least 1.5 to 1.55v for that, questions ist that safe for daily rig
i'm testing the settings now on memtest64, temps are fine with additional fan at 42c

edit on the pbo post:
i can lower my pbo allcore as much minus 25 but it wont change much on the volts, it went down from 1.4v allcore 4.6ghz to 1.35v at 4.65ghz allcore pbo from CO zero to CO minus 25allcore maybe the bios is or was bugged as i used F36e and this bios also caused the problem wih the vddp and vddg settings not sticking. i think i have to give it a try again on the older bios again. Are you using any offset? or just voltage on auto and curve optimizer

 

[Taraquin]

slightly better read speeds and more or less same latency, asking myself if i should use this setting as my daily or go for cl15 or even 14.. but as rares495 wrote i need at least 1.5 to 1.55v for that, questions ist that safe for daily rig
i'm testing the settings now on memtest64, temps are fine with additional fan at 42c

edit on the pbo post:
i can lower my pbo allcore as much minus 25 but it wont change much on the volts, it went down from 1.4v allcore 4.6ghz to 1.35v at 4.65ghz allcore pbo from CO zero to CO minus 25allcore maybe the bios is or was bugged as i used F36e and this bios also caused the problem wih the vddp and vddg settings not sticking. i think i have to give it a try again on the older bios again. Are you using any offset? or just voltage on auto and curve optimizer

I don't use offset, just plain +200 pbo and -30 on 4 cores and -29 on 2. I recently changed it to +50 and -30 allcore since I like silence and +200 PBO makes CPU cooler aggressive.

 

[schidddy]

I don't use offset, just plain +200 pbo and -30 on 4 cores and -29 on 2. I recently changed it to +50 and -30 allcore since I like silence and +200 PBO makes CPU cooler aggressive.

okay, first i run some tests for the mem to get it stable and use it some days on daily average tasks to avoid running into RAM problems while tuning on the CPU, then i try pbo again.
Any thoughts on running 1.55v on b-die DR for daily use?

 

[tabascosauz]

okay, first i run some tests for the mem to get it stable and use it some days on daily average tasks to avoid running into RAM problems while tuning on the CPU, then i try pbo again.
Any thoughts on running 1.55v on b-die DR for daily use?

1.55V is fine if you keep it cool. Which means a lot of airflow. Either you're running a hurricane through your case with extremely powerful fans, or have a push-pull downdraft cooler like me, or have dedicated ghetto fans for memory, or have your B-die waterblocked.

Why do you need 1.55V for 3800CL16 though? Even CL15 shouldn't need 1.55V.

No. they are on a2 and b2. It's weird need to test it without xmp to see if I got errors.

Turn on Geardown. Or drop to 2T.

You're never going to get anywhere by bruteforcing 1T no GDM with just VDIMM.

 

[schidddy]

1.55V is fine if you keep it cool. Which means a lot of airflow. Either you're running a hurricane through your case with extremely powerful fans, or have a push-pull downdraft cooler like me, or have dedicated ghetto fans for memory, or have your B-die waterblocked.

Why do you need 1.55V for 3800CL16 though? Even CL15 shouldn't need 1.55V.



Turn on Geardown. Or drop to 2T.

You're never going to get anywhere by bruteforcing 1T no GDM with just VDIMM.

i ment 1.55 for further experimentin lowering the timings, running the above setting at 1.45 now.
I got about 14fans in my Corsair 900D on Top 3x 140mm just above the RAMs and there is a little 60mm Fan on the grahpics card blowing directly on the DIMMs
no hurricane all silent keeping the water still on a very low delta

 

[Taraquin]

okay, first i run some tests for the mem to get it stable and use it some days on daily average tasks to avoid running into RAM problems while tuning on the CPU, then i try pbo again.
Any thoughts on running 1.55v on b-die DR for daily use?

If you have good cooling on them, a fan directly above that can be fine. Some airflow in your rig is nice to have.

No. they are on a2 and b2. It's weird need to test it without xmp to see if I got errors.

1T and no gdm often needs adjusting cad etc. 2T is my general advice. Gdm for those who like even numbers

 

[purecain]

T



This is what i have stable for now. I havnt tried overclocking the cores individually yet, still working out which are the best cores.

 

[Ibizadr]

Turn on Geardown.

You're never going to get anywhere by bruteforcing 1T no GDM with just VDIMM.

This is what I need to got it without errors. Thanks now it runs fine, it's time to some oc since this is a good kit

 

[Taraquin]

T



This is what i have stable for now. I havnt tried overclocking the cores individually yet, still working out which are the best cores.
View attachment 224404

What voltage on ram? If you are at 1.45V try 16 16 16 ras 32 rc 48 rrds 4 faw 16 wtrs 4 wtrl 10 rfc 272 rdrdscl 4 wrwrscl 4 rtp 8 rdwr 10 wrrd 3. If you are below 1.45V try increasing to 1.45V.

Procodt 43 or 48, 53 is too high. Set vddp to 880mv, 1098 is way too high.

 

[Ibizadr]

Hey guys thanks for your help now they run without errors on an 30min occt. I got it at 3800 CL18 but I think I can get it to CL16.
Here some ss of my timmings now and Aida benchmark. I will go to improve it and give updates

 

[Mussels]

okay @tabascosauz i guess i stay with the board, do you guys think i run into problems clocking the CPU to 4.8 to 4.9ghz if needed? i know it as an inferior VRM design but i am actually more likely to overclock efficiently then bruteforce method. Now im actually very fine with manual oc all@46 at 1.16v under load
@Taraquin ill try that this evening, thank you. Do you think i am in range with the RAM Volts with that values or do i need to set up? Its set to 1.45v hw info showing 1.47v

greetings

I think 4.7 is considered the ceiling for all core, with higher only achieved by PBO boosting lower core counts

If you can get a low voltage OC at 4.7, it's definitely worth considering. Seeing how i get 4.4-5.05 depending on PBO settings and amount of cores loaded, it comes down to whether you need that little bit of extra ST, or more MT performance.

 

[Taraquin]

Hey guys thanks for your help now they run without errors on an 30min occt. I got it at 3800 CL18 but I think I can get it to CL16.
Here some ss of my timmings now and Aida benchmark. I will go to improve it and give updates

Keep 1.45V dimm, try 16 16 16 ras 32 rc 48 rrds 4 rrdl 6 faw 16 wtrs 4 wtrl 8 or 10 rfc 280 or 272 wr 16 rtp 8 rdrd/wrwrscl 4 cwl 16 rest on auto

Set ccd volt and vddp to 900mv, iod to 1020mv. I'm quite sure that will work

These are lax timings that should work on most B-die.

 

[han32]

for daily only 1.45v and everything auto

 

[Dar1usElite]

Error 0 is usually too low voltage on ram. Try upping by 0.02V, usually fixes it. 15 can be several things, tRFC maybe, but that can often be improved by a bit more volt.

whats the maximum safe voltage i can go to? i don't want to go too high. ive heard 1.5v, ill go up by 0.02 and see how it goes

 

[Taraquin]

whats the maximum safe voltage i can go to? i don't want to go too high. ive heard 1.5v, ill go up by 0.02 and see how it goes

1.45V on B-die is considered totally safe. 1.5V is also considered safe, but you must have some airflow if you overclock\tweak a lot at 1.5V or it will overheat and get errors. B-die is very temp sensitive once tweaked\OCed and can throw error at below 50C. 50C can`t kill the ram so temp itself is not the issue.

 

[Dar1usElite]

1.45V on B-die is considered totally safe. 1.5V is also considered safe, but you must have some airflow if you overclock\tweak a lot at 1.5V or it will overheat and get errors. B-die is very temp sensitive once tweaked\OCed and can throw error at below 50C. 50C can`t kill the ram so temp itself is not the issue.

Alright cool its running atm. It spat 2 errors at me but again I didn't launch it in admin mode, unsure why this happens but ‍. Dram temps are at about 55c according to hwinfo.
This is hynix djr though
Edit :I went out and came back and it's currently on its 8th run and given 1 error with code 13

 

[Taraquin]

Alright cool its running atm. It spat 2 errors at me but again I didn't launch it in admin mode, unsure why this happens but ‍. Dram temps are at about 55c according to hwinfo.
This is hynix djr though
Edit :I went out and came back and it's currently on its 8th run and given 1 error with code 13

I'm not sure how DJR handles heat. I know B-die and D-die is sensitive, rev E don't care. Post a zentimings screenshot? Error 13 can be too little voltage.

 

[Dar1usElite]

I'm not sure how DJR handles heat. I know B-die and D-die is sensitive, rev E don't care. Post a zentimings screenshot? Error 13 can be too little voltage.

 

[Taraquin]

View attachment 224510

Did you get less errors after increasing voltage? Try upping to 1.48V.

 

[Dar1usElite]

Did you get less errors after increasing voltage? Try upping to 1.48V.

i think i only got the 1 error code 13 over 11 runs, previously i had 2. I was a little concerned though about the ram temps as one dimm was at 59c

 

[Taraquin]

i think i only got the 1 error code 13 over 11 runs, previously i had 2. I was a little concerned though about the ram temps as one dimm was at 59c

Try running 1.42V, see if that does the same? Maybe just one timing is a bit off.

 

[Dar1usElite]

Try running 1.42V, see if that does the same? Maybe just one timing is a bit off.

so i ran it overnight with 20 runs. got 8 errors, i forgot to screenshot it but most were error 0, one error 5 and an error 14 that i seen. Max ram temp was 54c
going to retry at 1.47v with a fan on the ram to see if its temp related

so i ran it overnight with 20 runs. got 8 errors, i forgot to screenshot it but most were error 0, one error 5 and an error 14 that i seen. Max ram temp was 54c
going to retry at 1.47v with a fan on the ram to see if its temp related

@Taraquin so i reran the test twice, once with no fan on the ram and once with a fan on it. (attached is a photo of the run with the fan OFF).
1.47v both times.
second run got 1 error 4 and that was it so it seems heat makes it worse but doesn't solve it.

not sure why it added to the last post but yeah.