Overclock.net banner

Viper Steel 4000 16-16-16-36 overclocking for gaming

8.9K views 10 replies 9 participants last post by  Ichirou  
#1 ·
MOBO: MSI PRO Z690-A
CPU: 12600KF E-CORE DISABLED P CORE 5.0 RING 4.5 1.35v
RAM: VIPER STEEL PVS416G400C6K

Hi there,
Im lookin for advice on overclocking this bdie kit for gaming, not trying to break any records.

This is where im at atm:
Image

tXp: 6
PPD: 0
DRAM: 1.48v
SA: 1.3v

All the help is very much appreciated!
 
#2 · (Edited)
set ram to 1.54 volts
set ram speed to 4100
try 16-16-16-35
bring tREFI down to ~35000
twr can probably go down to 12
 
#3 · (Edited)
MOBO: MSI PRO Z690-A
CPU: 12600KF E-CORE DISABLED P CORE 5.0 RING 4.5 1.35v
RAM: VIPER STEEL PVS416G400C6K

Hi there,
Im lookin for advice on overclocking this bdie kit for gaming, not trying to break any records.

This is where im at atm:
View attachment 2568574
tXp: 6
PPD: 0
DRAM: 1.48v
SA: 1.3v

All the help is very much appreciated!
DRAM 1.53v \ 1.55v
SA 1.3v
VDDQ 1.2v \ 1.25v
tXP = Auto

Image
 
#8 · (Edited)
DRAM 1.53v \ 1.55v
SA 1.3v
VDDQ 1.2v \ 1.25v
tXP = Auto

View attachment 2568588
Hey,
Ive tested ur timings with dram 1.55 and vddq 1.25, but:
Image


set ram to 1.54 volts
set ram speed to 4100
try 16-16-16-35
bring tREFI down to ~35000
twr can probably go down to 12
Cheers, will try tomorrow, should i keep the other timings the same? Yes, tWR can go to 12, ive already tested.

Try tREFI of 65534 (for performance increase)

When you're done optimizing the rest, try lowering tRFC to somewhere around 240.

If cooling allows (it probably does), try DRAM voltage of 1.58 (unless contraindicated for 12th gen - I'm not sure)

try RDWR 12, 12, 12, 13
RDRD_sg of 6
WRWR_sg of 6

tWR of 14, 12, 10, then 8

Try tWTR_L of 8
and tWTR_s of 4

edit: tCWL of 14, then 12

I have the 4400MHz kit (basically the same thing) and these are some of my timings. Where I listed many it's the lowest

WRRD_sg and WRRD_dg are used to lower tWTR_L and tWTR_s (you lower WRRD_sg to lower tWTR_L and WRRD_dg to lower tWTR_s)
Looks like lowering tRFC decreases FPS on any 12th gen except i9, ive heard this a couple of times. Personally going lower than 320 tRFC decreases FPS.
 
#5 · (Edited)
Try tREFI of 65534 (for performance increase)

When you're done optimizing the rest, try lowering tRFC to somewhere around 240.

If cooling allows (it probably does), try DRAM voltage of 1.58 (unless contraindicated for 12th gen - I'm not sure)

try RDWR 12, 12, 12, 13
RDRD_sg of 6
WRWR_sg of 6

tWR of 14, 12, 10, then 8

Try tWTR_L of 8
and tWTR_s of 4

edit: tCWL of 14, then 12

I have the 4400MHz kit (basically the same thing) and these are some of my timings. Where I listed many it's the lowest

WRRD_sg and WRRD_dg are used to lower tWTR_L and tWTR_s (you lower WRRD_sg to lower tWTR_L and WRRD_dg to lower tWTR_s)
 
  • Rep+
Reactions: storm-chaser
#7 ·
Hmm, maybe my thinking that DRAM voltage of 1.58 is possibly too high for Intel's 12th gen could be wrong.
I have seen what I consider to be ridiculously high SA voltages now with Intel's new(ish) 12th gen processors - over 1.5V!

Does anyone know what's up with this?

My thinking is CPU transistors are very small, but a fraction of a volt shouldn't make a difference to the longevity of a transistor, no matter it's size. If the resulting current is too high? You'll get damage. If the heat is too high? Damage. But a small fraction of supply voltage? It's unlikely to cause any problem at all.
Unfortunately, they don't have a temperature sensor to measure the temperature of the transistors used by the CPU to interface with the RAM. Maybe they should put one.

Intel??

Then again, maybe they don't because the number of transistors is too small to effectively measure (hot-spots)
 
#9 ·
My board doesn't even let me set them to 1 as a minimum, 4 is the lowest? I just have them all set to the standard 7's, and 12's on tRDWR dr,dd. everyone is posting in DDR4 12th gen thread here. Totally maxed out the tREFI 65535. tWRRD sg,dg could be a bit lower 26,23.
Anyway I just tested tRFC between 270-320 on Shadow of tomb raider bench and didn't see difference (within margin of error) it was all averaging 122 FPS in the test not 180+ FPS with your call of duty which could be the difference.