Ryzen 9 3900x Voltage and Temperature

[azngreentea01]

Doing some benchmark with Cinebench R20 . I notice something funny is when the computer is idle the voltage at 1.45-1.46 V and temp is around 45-55 Cel.

When i started doing the benchmark, the Voltage drop to 1.3- 1.4 v. The CPU core is at 4.0 ghz on all cores.

 

[lexluthermiester]

Doing some benchmark with Cinebench R20 . I notice something funny is when the computer is idle the voltage at 1.45-1.46 V and temp is around 45-55 Cel.
View attachment 126650
View attachment 126651View attachment 126652

When i started doing the benchmark, the Voltage drop to 1.3- 1.4 v. The CPU core is at 4.0 ghz on all cores.

View attachment 126653

That voltage is WAY too high! For first gen Ryzen 1.35 to 1.4 was common, but these new gen Ryzen CPU's should only be at 1.2 or 1.25.

See this and start at 3:32;

Jay will walk you through what to set and why.

 

[HenrySomeone]

Ah, yes, the delights of AMD ownership...

 

[azngreentea01]

So, what I did was. I offset my cpu by - .1 in the bios. Now it showing 1.35 - 1.39 v

After rerun the benchmark. It showing better result.

 

[lexluthermiester]

Ah, yes, the delights of AMD ownership...

Actually, the problems here are not the fault of AMD, it's the board maker not setting the voltages in the BIOS correctly based on the CPU microcode

So, what I did was. I offset my cpu by - .1 in the bios. Now it showing 1.35 - 1.39 v
View attachment 126660

After rerun the benchmark. It showing better result.

View attachment 126661

Are your temps better as well? I'd still set the voltage lower, say -0.2, if you're not OCing.

 

[Space Lynx]

That voltage is WAY too high! For first gen Ryzen 1.35 to 1.4 was common, but these new gen Ryzen CPU's should only be at 1.2 or 1.25.

See this and start at 3:32;

Jay will walk you through what to set and why.

Jay is great. Always enjoyed his videos, don't watch anyone too much these days, except Linus once in awhile because he is goofy as crap sometimes. lol

 

[azngreentea01]

Actually, the problems here are not the fault of AMD, it's the board maker not setting the voltages in the BIOS correctly based on the CPU microcode

Are your temps better as well? I'd still set the voltage lower, say -0.2, if you're not OCing.

Yes temp drop down to 38-45 Cel. I think i stay at this offset at -0.1 for now and let see on gaming. Voltage will drop down 0.850v when idle and when is use is at 1.345 V, No i don't think i will OCing because of the state we in right now. lots of problem with the motherboard bios.

 

[Space Lynx]

Actually, the problems here are not the fault of AMD, it's the board maker not setting the voltages in the BIOS correctly based on the CPU microcode

Are your temps better as well? I'd still set the voltage lower, say -0.2, if you're not OCing.

Honestly Asus has always done this. Even on my Asus Z68 Deluxe or w.e it was called, whenever I flipped the switch to auto OC it on the mobo, called TPU and EPU, it would skyrocket high voltage and only give me 4.3ghz oc on sandy bridge 2500k... everyone always just manually oc'd a 4.5 clock at like much lower voltage than that.

Asus is just overrated as a company as far as I am concerned.

 

[oxrufiioxo]

Similar issues on my Z390 Asus Maximus Code... Regardless of bios version the stock voltages for vccio and SA are way too high.

 

[TheMadDutchDude]

All vendors seem to do it. I asked MSI (on the X99 platform) why this was... and their best response was: "it provides guaranteed stability."

Yeah... 1.35 VCCIO on a CPU that only needs 1.05v to function is not what I would deem safe, nor stable. Not to mention that anything over 1.3v on air/water was actually considered a damaging prospect too.

 

[Metroid]

That is something that i dont like, if cooling is at 70c, why 4.0ghz? supposed to be all at 4.6ghz, i wonder when the cpu starts throttling, 60c? does anybody know?

What the hell this is supposed to work all with 4.6ghz? 30c and vcore 1.2? well as cold as it gets vcore can be dropped little by little and maintains maximum clock speed allowed 4.6 on 3900x, or same can be said about as hot as it gets vcore and clock speed lowers to maintain temperature that is good to make that vcore stable.

Here at the techpowerup boos analysis, wizzard forgot to show the current temperature graph scale too. I wonder if temperature less than said 50c will still leave it at 4.2ghz maximum on all cores. LN2 bench can confirm my theory or a very good watercooling.

AMD Ryzen 9 3900X Review

The flagship of AMD's new Ryzen 3000 lineup is the Ryzen 9 3900X, which is a 12-core, 24-thread monster. Never before have we seen such power on a desktop platform. Priced at $500, this processor is very strong competition for Intel's Core i9-9900, which only has eight cores.

www.techpowerup.com

 

[harm9963]

Waiting till Christmas , lots BIOS updates till then.
learn the hard way with GD80 990FXA,BIOS updates every month for 8 months back in 2011.

 

[Metroid]

Doing some benchmark with Cinebench R20 . I notice something funny is when the computer is idle the voltage at 1.45-1.46 V and temp is around 45-55 Cel.

When i started doing the benchmark, the Voltage drop to 1.3- 1.4 v. The CPU core is at 4.0 ghz on all cores.

Can you disable smt and see if you can overclock it better, also can you see power consumption with smt off, 100% load? I just don't understand why no reviewers have done this test.

 

[ShurikN]

No i don't think i will OCing because of the state we in right now. lots of problem with the motherboard bios.

You wont get much from overclocking anyway, don't bother. PBO and XFR do a good job. In some cases (like lightly threaded games) you might even lose performance versus stock.

 

[racer243l]

All vendors seem to do it. I asked MSI (on the X99 platform) why this was... and their best response was: "it provides guaranteed stability."

Yeah... 1.35 VCCIO on a CPU that only needs 1.05v to function is not what I would deem safe, nor stable. Not to mention that anything over 1.3v on air/water was actually considered a damaging prospect too.

My Asus Maximus XI Hero used 1.45v VCCIO and 1.4v VCCSA on 3600Mhz XMP and almost 1.38v on the CPU for 4.7Ghz allcore turbo
Manually I got down to 1.365v VCCIO and 1.34v VCCSA with 4000Mhz RAM and 1.3v for 5Ghz on the CPU.

 

[the54thvoid]

I can't get it now (on mobile) but I saw a Reddit thread from an AMD rep, could even have been a guy called Rob. But... Monitoring software isn't reporting idle voltage. Effectively, the software wakes the core, which responds to the call as an intense load. Therefore, the idle voltage is actually a load voltage, instigated by the monitoring. Ultimately, it's like quantum. The act of observing the chip, changes it's state from cc6.
Only one single software or bios monitoring tool should be active. I think he said either CPUz was okay or Ryzen master. But only one. Mobo monitoring tools will mess either up.

 

[Hardi]

I can't get it now (on mobile) but I saw a Reddit thread from an AMD rep, could even have been a guy called Rob. But... Monitoring software isn't reporting idle voltage. Effectively, the software wakes the core, which responds to the call as an intense load. Therefore, the idle voltage is actually a load voltage, instigated by the monitoring. Ultimately, it's like quantum. The act of observing the chip, changes it's state from cc6.
Only one single software or bios monitoring tool should be active. I think he said either CPUz was okay or Ryzen master. But only one. Mobo monitoring tools will mess either up.

https://www.reddit.com/r/Amd/comments/cbls9g

 

[Vlada011]

Finally day when gamers could buy killer PC for 2000$.
3900X, 1TB Gen 4 M.2-Radeon 5700XT CH4H Motherboard and 16GB RAM.

 

[HenrySomeone]

A killer 2k$ PC with a Radeon 5700XT? Current crap blower ones no less?

 

[HD64G]

Some BIOS and driver tuning is still needed to allow all CPUs to perform as intended for the boost frequencies.

 

[the54thvoid]

Now a TPU news post.

 

[azngreentea01]

I can't get it now (on mobile) but I saw a Reddit thread from an AMD rep, could even have been a guy called Rob. But... Monitoring software isn't reporting idle voltage. Effectively, the software wakes the core, which responds to the call as an intense load. Therefore, the idle voltage is actually a load voltage, instigated by the monitoring. Ultimately, it's like quantum. The act of observing the chip, changes it's state from cc6.
Only one single software or bios monitoring tool should be active. I think he said either CPUz was okay or Ryzen master. But only one. Mobo monitoring tools will mess either up.

I undestand the observer effect on TPU article, but why is it the Temp is so high. I mean if it in idle the Temp should be around 35-45 Cel. When it idle it sitting around 45-55 cel.
I am only running one instance monitor tool. Not opening up 2-3 tools to monitor.


Edit:
ALso, when i use or turn on Icue software from Corsair. The voltage will be at contanst 1.35- 1.37V when idol . When i turn it off the idol will drop down to .885v.

 

[mngew]

Doing some benchmark with Cinebench R20 . I notice something funny is when the computer is idle the voltage at 1.45-1.46 V and temp is around 45-55 Cel.
View attachment 126658
View attachment 126650
View attachment 126651View attachment 126652

When i started doing the benchmark, the Voltage drop to 1.3- 1.4 v. The CPU core is at 4.0 ghz on all cores.

View attachment 126653

My 3900X also idles at those temperatures. I have mine cooled by H115i. I reseated the heatsink 3 times. Didn't make any changes. I'm going to try NH-D15 later tonight.

 

[lexluthermiester]

My 3900X also idles at those temperatures. I have mine cooled by H115i. I reseated the heatsink 3 times. Didn't make any changes. I'm going to try NH-D15 later tonight.

Like him, check your voltages in the BIOS. If they're above 1.25, set them to 1.25.

 

[Prillan]

Old post but i found that with gaming mode turned on in dragon center will push the cpu all the time hence the extreme idle temps, went from 45-50c to about 35-40 when i turned it off, maybe different utility programs are messing with eachother