[SOLVED] define cpu voltage and ghz on ryzen 5 2600

bradur

Commendable
Jan 15, 2019
4
0
1,510
So i want to overclock my ryzen 2600 to the highes ghz with the lowest voltage and minimal damage done but im a newby in oc and i dont know anything. after researching i still didnt understand how to figure out your voltage or when its seriously too much.

so my specs are:
Ryzen 5 2600 Be quiet pure rocket slim
Msi b450 gaming plus>
lpx vengeance 16gb ddr4 2666
gtx 970

my cpu is currently running at 3.7ghz and 1.225 voltage with aida64 stresstest max 85 temp

Sorry for my bad english i dont usually wrtite forum posts :)


 
Solution
Keep voltage at 1.35 and work your way up, most of them can comfortably do 3.9-4.0 at that voltage

Once you find your max oc you can slowly walk voltage back down to help with temps if you want

bradur

Commendable
Jan 15, 2019
4
0
1,510


so should i just put the voltage back to 1.35 and then continue as you said? or slowy work it up

and what temps do i have to look out for and where should i stop?
 
https://www.gamersnexus.net/guides/3251-raven-ridge-soc-voltage-guidelines-how-to-kill-cpu-with-safe-voltage

Watch the video and read the article before doing anything, or at bare minimum the conclusion including:

"If you ask us, we think decent overclocks are achievable with 1.2-1.25V SOC going into the socket, and that more than that isn’t really necessary, anyway."

Steve's a pretty smart guy... also for every 100MHz you increase for Ryzen beyond roughly 3.7GHz the voltage requirement (and thus temperature) goes up significantly. It's a very, very sharp ramp up.

Voltage increases the temperature but a high voltage at acceptable temperature can still kill your CPU.
 
Update: since it's not immediately clear, know that "Raven Ridge" is Zen+/2000 series. The 1000 series is Summit Ridge.

And 1.35V SOC will most likely KILL YOUR CPU (or at least degrade) over time.

Update 2: my bad, I guess RAVEN RIDGE is just the mobile and some desktop CPU's/APU's but the R5-2600 is not "Raven Ridge" so I'm not sure if the SOC values for just the CPU (not iGPU) are exactly the same as in the above video... Steve is testing a DESKTOP APU

It's also very confusing because some sites like Tweak Town are suggesting voltages that Gamers Nexus (and others) say will kill the CPU: https://www.tweaktown.com/guides/8547/amd-raven-ridge-2000-series-zen-apu-overclocking-guide/index3.html
(they recommend around "1.2V" but then also say don't go over 1.45V... WTF?)

*Anyway, for now I'm going to assume the safe CPU SOC voltage for Zen+ and desktop Raven Ridge are the same. Unless someone can find a good link that says the voltages are different.

("SOC" means System on a chip... meaning the CPU and GPU etc so I guess a regular CPU with no iGPU should not have an SOC value though the CPU voltage should be equivalent to the Raven Ridge video for "SOC" which is applied to the CPU and would be even if you disabled the iGPU)

Sorry for the confusion.
 

bradur

Commendable
Jan 15, 2019
4
0
1,510


so i watched the video and oc to 3.9 with 1.25v and auto socv at 1.0-1.1

i guess my mainboard can handle that so my cpu doesnt die

 

davew1860

Prominent
Feb 6, 2019
145
29
620


so should i just put the voltage back to 1.35 and then continue as you said? or slowy work it up

and what temps do i have to look out for and where should i stop?

How are you doing your thermal testing?

If using Prime95 "Small FFT's" for 15 minutes, try to keep your temperature under 80 C.
 
  • Like
Reactions: TJ Hooker

slavi_asenov2002

Honorable
Apr 11, 2018
117
6
10,695
Here is an excellent and clear CPU overclocking guide written by one of Tom's Hardware's very own: https://forums.tomshardware.com/faq/cpu-overclocking-guide-and-tutorial-for-beginners.3347428/

Not only should you slowly adjust voltage. You should also slowly adjust CPU speed. Increasing the speed too quickly can lead to crashes.
I read the article,. Nice brief explanation but it takes a lot of time. In my opinion, I would go for max 30 minutes of prime95 testing, if the cpu frequency is not high as I want.
For example my goal is 4.2 ghz so I try to do the tests for 4.0/4.1 faster and when I reach 4.2ghz with a somewhat stable voltage I do a 4-8 h test, then I try to lower the voltage.
I have done it with my cpu but I hit a wall for 4.1 ghz and I decided I will play with loading the voltage for 4.0 ghz and at he end from 1.35 volts I managed to lower it to 1.2375v.

A guide for prime 95 testing - https://overclocking.guide/stability-testing-with-prime-95/
 
  • Like
Reactions: davew1860

davew1860

Prominent
Feb 6, 2019
145
29
620
I read the article,. Nice brief explanation but it takes a lot of time. In my opinion, I would go for max 30 minutes of prime95 testing, if the cpu frequency is not high as I want.
For example my goal is 4.2 ghz so I try to do the tests for 4.0/4.1 faster and when I reach 4.2ghz with a somewhat stable voltage I do a 4-8 h test, then I try to lower the voltage.
I have done it with my cpu but I hit a wall for 4.1 ghz and I decided I will play with loading the voltage for 4.0 ghz and at he end from 1.35 volts I managed to lower it to 1.2375v.

A guide for prime 95 testing - https://overclocking.guide/stability-testing-with-prime-95/

Yeah, that is a safe and more efficient way to reach an overclock goal. You can quickly weed out the overambitious CPU speeds that way.

Under any circumstances, I personally wouldn't run the "Small FFTs" torture test for more than 15 minutes at a time because it so dramatically increases one's temperatures.