My Ryzen 7 3700X OC.

[freeagent]

Your CPU has a default max PPT of 88W, EDC of 90A and TDC of 60A.
Now in that screenshot you have
PPT: 115W
EDC: 125A
TDC: 70A

The actual power consumption (PPT) of 115W is not all that bad, even if it was 125 or 150W...
The TDC of 70A is not all that significant IF you keep an eye on temps. TDC is "Thermal Design Current" means the max allowed Current (A) under thermal throttling if and when that occurs (>= 95C). And it should be over 60A.

The most concerning is EDC (ElectricDesignCurrent) value of 125A. 7nm process node has thinner traces and transistor gates. The worst enemy of any chip as an integrated circuit is EMI (electromigration).

"Electromigration decreases the reliability of chips (integrated circuits (ICs)). It can cause the eventual loss of connections or failure of a circuit."

"With increasing miniaturization, the probability of failure due to electromigration increases in VLSI and ULSI circuits because both the power density and the current density increase. Specifically, line widths will continue to decrease over time, as will wire cross-sectional areas. Currents are also reduced due to lower supply voltages and shrinking gate capacitances. However, as current reduction is constrained by increasing frequencies, the more marked decrease in cross-sectional areas (compared to current reduction) will give rise to increased current densities in ICs going forward."

We are not pulling stuff out of our arse because we dont like any one to have their fan or to not OC just because we like it that way... The risks are real and 7nm is too new for users and past experience with 12/14nm or grater nodes is completely irrelevant.
From my understanding what you (sadly and unfortunately) fail to grasp is that you cant use that voltage with any load/current. It seem wierd and odd to you why we suggest that 1.35~1.4V is too much when stock settings are pushing even 1.5V, right? I've said it a few times but you seem to skip it or dont want to understand it, I really dont know...
Stock settings are supplying high voltages when load/current is low only! If load gradually increases, up to max, the voltage is gradually dropped to 1.2V~1.3V (depends if its AVX or not) to keep current under check. The voltage and speed is determined by the quality of chip's silicon but its around that.

The major cause of EMI is high current.-
High temperature is worsen it by far, even if its under throttle temp (95C). For once more... the videos I provided show a way to determine the safe voltage. Because the CPU it self when on stock is monitoring current and temp to regulate speed and voltage to preserve silicon. These CPUs have an internal silicon manager/controller called FIT. (Silicon FITness controller).
If you had better cooler and your temp was like 70~75C then a 125A wouldnt be that bad... At 80+C is a very different story.

By the author of HWiNFO:
CPU "Core VIDs" are just a requested voltage of each core and not the actual supplied voltage. For starters the VRM system cannot supply multiple voltages but only one. The "CPU Core VID (effective)" is the "winning" request among all, but still is not the supplied voltage.
"Vcore" is the CPU core voltage reading by the board sensor (could by inaccurate and probably is).
"VR VOUT" (if exists) is the voltage reading by the VRM sensor (could by inaccurate and probably is, even on high quality VRMs).
"CPU Core Voltage (SVI2 TFN)" is a reading value pulled straight from inside the CPU and most accurate core voltage and closest to real as possible.

Again... I'm not the one saying this but the person who wrote HWiNFO.


You can use DRAM XMP/DOCP without any fear of damage. The worst case scenario is to be unstable but damage is out of the question. If its unstable the you could try to make it stable, with some help of course.



Its fine! The only "wrong" thing I can see is the low DRAM speed and the asynchronous speeds of UCLK and FCLK. The system is missing some performance.

Thank you for taking the time to write this out man! Honestly.. I don't really notice stock vs oc anyways. I don't have PBO or core enhancement on and it hasn't been stabbed with 1.5v yet. My kids were using the system to play on Steam.

I do have to agree with Trickson.. it is fast af as it is. I might just buy an Intel system I can abuse.

 

[trickson]

If it's stock with either/both boost turned on that's fine, unless you're forcing it all core overclock at that voltage.

The RAM clock speed is 2133 MHz, might want to turn on the memory profile A-XMP in the BIOS.

Okay what are you trying to say that the AMD CPU can not be overclocked due to some voltage crap? Man stop scaring people!
Your voltage is FINE if you do not play with it and leave it on auto AMD and your MB should handle things fine and yeah force the 4.4GHz why NOT? If you keep the CPU cool it will be fine at 1.4Volts I have NOT seen ANY proff that 1.4 Volts will in any way kill the AMD Ryzen CPU! TO DATE! NOT A SHRED!
Stop telling people lies! AMD Has ALWAYS used 1.4 and even 1.45 as a mark as well as Intel there CPU's can handle and should be able to take 1.4+ for heavy work and gaming ! I have not seen any issues with 1.4 and 1.45 in fact if I let the CPU take control I can hit 4.5 GHZ no problem but I do not need that 4.4 is GREAT!

PS I am not calling biffzinker a liar. OKAY

I am just saying in general if you have NO proof that the AMD CPU Ryzen LINE is dyeing because of voltages higher than 1.35 then you best have some hard ass evidence and not some bullshit Youtube dummy crap! because that is a not going to fly with me NO SIR NOT AT ALL!
The ONLY thing I have found with the NEW Ryzen line of CPU's is the Memory controller, In the first gen they suck can not get RAM to run at ratted speed best you get is 1066MHz! WTF! Other than that the CPU's are KICKASS and can OC fine and they can take the Voltage just fine ( Proof is still pending).

 

[biffzinker]

It's a perfect OC. I knew it NO one can give us any ANY real information on this Voltage issue It would seem you are all going off some Youtube dummy.
Very little information if you ask me. That is NOT proof that MY CPU will get some BS degradation that is such BS IMHO anyway!

My Ryzen 5 3600 went kaput either from me trying an overclock like you are or from the earlier AGSEA firmwares when AMD was working on the firmware. I do know I tried a forced all core overclock at a high voltage, then later did a PBO overclock that was stable that then started causing BSOD in Windows, and then could no longer set the PBO overclock from before without it crashing on boot.

My 3600 went from 4.3 GHz stable at 1.37 down to 4.050 at 1.28

 

[trickson]

My Ryzen 5 3600 went kaput either from me trying an overclock like you are or from the earlier AGSEA firmwares when AMD was working on the firmware. I do know I tried a forced all core overclock at a high voltage, then later did a PBO overclock that was stable that then started causing BSOD in Windows, and then could no longer set the PBO overclock from before without it crashing on boot.

My 3600 went from 4.3 GHz stable at 1.37 down to 4.050 at 1.28

And the Mother Board you were / are using?
A poor MB with Substandard parts can cause issues.
I had enough of MSI MB's the quality has been just awful I was able to get the R7 1700x to 3.8GHz stable but it bricks out at 4.0GHz Now no matter the voltage. It's NOT the CPU as the CPU was fine at 4.0GHz the MB on the other hand has to be the issue! The Gigabyte MB has NO problem taking the R7 1700X to 4.0GHZ not at all. But that is NOT the MB for that CPU as you can see in sig. I have tried mine in another system and found this to be the case, MSI SUCK ASS! Try another MB! I should have learned from the FX8300 that is FINE just the MB. Have you tried it in a different MB?
No one ever talks about this part of the equation they use cheep ass MB's and claim that is the CPU did it. Not ever considering It could be some POS MOSFET or CAP on the MB! Power is very important and it would seem ASUS did it right for all Ryzen CPU's!

 

[biffzinker]

Same board as in my System Specs, swapping to the 3800X solved the no CPU detected (CPU LED was lit.) All I did was a normal shutdown that night, next morning it wasn't detecting the 3600. The B450 Tomahawk isn't a cheap sub-standard board even for MSI.

Unwatching this thread. Good luck

 

[trickson]

Same board as in my System Specs, swapping to the 3800X solved the no CPU detected (CPU LED was lit.) All I did was a normal shutdown that night, next morning it wasn't detecting the 3600. The B450 Tomahawk isn't a cheap sub-standard board even for MSI.

Yeah that sounds just like an MSI mb for you! Get a real MB and try then.

 

[RJARRRPCGP]

Just got my internet back Comcrap once again shut us down without any notice for 6 hours so nice of them!

Yikes! I never had that happen that I was aware of, but when I lived in Bellows Falls, Vermont, which was from June 12, 2016 to February 24, 2018, when I had Comcast, I would get randomly get disconnected! I would suffer random outages and they seemed more likely during night time!

Yeah that sounds just like an MSI mb for you! Get a real MB and try then.

Really?! I know that was true back in the "new-FX-era"!

Thank you for taking the time to write this out man! Honestly.. I don't really notice stock vs oc anyways. I don't have PBO or core enhancement on and it hasn't been stabbed with 1.5v yet. My kids were using the system to play on Steam.

I do have to agree with Trickson.. it is fast af as it is. I might just buy an Intel system I can abuse.

I already did buy a Comet Lake build! In case I can't get a Vermeer, even with months passed!

 

[freeagent]

I trust many opinions here, and I don’t have the experience to say otherwise. But thanks to the fellas for helping us both, always learning!

 

[Deleted member 202104]

Reading this thread is like watching "Faces of Meth" in real-time.

 

[RJARRRPCGP]

The B450 Tomahawk isn't a cheap sub-standard board even for MSI.

Which BIOS? 1E? I honestly appeared to have better Cinebench results with 1E, when using a Matisse, IIRC. The original BIOS, was 1C, IIRC.

 

[trickson]

Same board as in my System Specs, swapping to the 3800X solved the no CPU detected (CPU LED was lit.) All I did was a normal shutdown that night, next morning it wasn't detecting the 3600. The B450 Tomahawk isn't a cheap sub-standard board even for MSI.

Unwatching this thread. Good luck

So you leave if confronted with a question? Or because some one is questioning your results or the bull crap about voltage (Because that is how I see it Bull CRAP)?
I am confused are you mad? I hope not this is NOT about YOU.

 

[Toothless]

Because my request to be locked from this thread hasn't been given, here we go again to get through how dense you are.

Ryzen 3000 Safe Voltages and Degradation

Hello, I guess you know where this is going... So, I finally moved to a 3800x build! I love it, it is awesome. Tuned 3800mhz RAM at 1:1:1 ratio, beautiful! I also managed to tune all cores to 4.425ghz @1.325V. I’m using a Noctua NH-u14s with two fans. Temps running CB20 loop or AIDA stress are...

community.amd.com

Last year, when I was reading about the TSMC 7nm node they (TSMC) were quoted as saying that it was specified for 1.3 Volts and that is what I have kept to.

AMD has a vested interest in declaring higher voltages "Safe" because their Marketdroids want to plaster as high a boost clock onto their junk as possible. And they don't give a damn as long as the CPU will at least work for the duration of the warranty.

Wow, on our own forum, bet you never cared to see this in your "research"

Ryzen 3000 degradation with static voltage and static frequency

Hi im new, and english is not my native language but ill doing my best without a translator. im here because ive been reading about"safe voltages"overclock, and that stuff direclty on the bios. I understand that but my thought is the next Ive been reading that higher voltages can hurt, lower...

www.techpowerup.com

Discussion - Ryzen 3000 CPUs overclocking: exercise caution

Buildzoid has admitted that the static overclock of his AMD Ryzen 7 3700x has degraded the processor. More on it here: https://www.youtube.com/post/UgwIDprhUtrJzFuhW0N4AaABCQ

forums.anandtech.com

They link straight to a comment, that confirms one case of a chip degrading. If you're going to dismiss one of the better analysts of boards (wow, something you're so anal about) then your headache is more than just that.

Actually Hardcore Overclocking

Ok so it looks like I've managed to degrade my 3700X as it crashes on MSI mobos even at stock and PBO doesn't work on it at all now. EDIT I think since I alr...

www.youtube.com

what vcore do you guys run 24/7 ryzen 3000s

So got my system stable with a all core clock of 4.4ghz on 1.315 drops down to 1.28 under load have tossed 6-8hrs of realbench at it 4 hrs of cb20 4-6 hrs of aida and gamed and i run metatrader also its nice and stable i dont really wanna go any higher on vcore about as high as i feel safe...

hardforum.com

Here we go over the safe max of 1.325v, which in other news was dropped to 1.3v to being safe. Still don't accept the reality? It's okay, we know you're dense.

https://forums.whirlpool.net.au/archive/3vy0pnp9

Nifty, more talk about the voltage issues. It's almost like AMD is smarter than you about what their chip needs. Let me go over that again.

Idle, the chip goes to between 1.4v-1.5v, probably for stability reasons. This is a low current load. Notice low current?

Under load, the chip drops to below 1.3v and in range of safe current amounts in the processor. Wow, it's like the engineers there know something.

Hang on, lemme get this to you.

TSMC spec at 7nm for 1.3V

Is that big enough for you? It's like you sit on a forum and say "screw all of you" when this is your first 7nm chip. You sit there all high and mighty with what? 14nm? 12? Since when did you make chips better than TSMC and call forward what THEIR chips need?

Again, the difference is the VOLTAGE UNDER LOAD compared to idle. These are not 2700x or older. These are not Intel. Hell at this point you should've gone with a 9900k since Intel can't get their die production fixed after how many years.

Oh, and you're probably having boot issues because of your ram, IF, and something else not being tweaked because "moar voltz moar hertz" at this point. I might have to get myself locked pretty hard from the thread since it keeps popping back up in my feed.

 

[tabascosauz]

Because my request to be locked from this thread hasn't been given, here we go again to get through how dense you are.

Ryzen 3000 Safe Voltages and Degradation

Hello, I guess you know where this is going... So, I finally moved to a 3800x build! I love it, it is awesome. Tuned 3800mhz RAM at 1:1:1 ratio, beautiful! I also managed to tune all cores to 4.425ghz @1.325V. I’m using a Noctua NH-u14s with two fans. Temps running CB20 loop or AIDA stress are...

community.amd.com

Wow, on our own forum, bet you never cared to see this in your "research"

Ryzen 3000 degradation with static voltage and static frequency

Hi im new, and english is not my native language but ill doing my best without a translator. im here because ive been reading about"safe voltages"overclock, and that stuff direclty on the bios. I understand that but my thought is the next Ive been reading that higher voltages can hurt, lower...

www.techpowerup.com

Discussion - Ryzen 3000 CPUs overclocking: exercise caution

Buildzoid has admitted that the static overclock of his AMD Ryzen 7 3700x has degraded the processor. More on it here: https://www.youtube.com/post/UgwIDprhUtrJzFuhW0N4AaABCQ

forums.anandtech.com

They link straight to a comment, that confirms one case of a chip degrading. If you're going to dismiss one of the better analysts of boards (wow, something you're so anal about) then your headache is more than just that.

Actually Hardcore Overclocking

Ok so it looks like I've managed to degrade my 3700X as it crashes on MSI mobos even at stock and PBO doesn't work on it at all now. EDIT I think since I alr...

www.youtube.com

what vcore do you guys run 24/7 ryzen 3000s

So got my system stable with a all core clock of 4.4ghz on 1.315 drops down to 1.28 under load have tossed 6-8hrs of realbench at it 4 hrs of cb20 4-6 hrs of aida and gamed and i run metatrader also its nice and stable i dont really wanna go any higher on vcore about as high as i feel safe...

hardforum.com

Here we go over the safe max of 1.325v, which in other news was dropped to 1.3v to being safe. Still don't accept the reality? It's okay, we know you're dense.

https://forums.whirlpool.net.au/archive/3vy0pnp9

Nifty, more talk about the voltage issues. It's almost like AMD is smarter than you about what their chip needs. Let me go over that again.

Idle, the chip goes to between 1.4v-1.5v, probably for stability reasons. This is a low current load. Notice low current?

Under load, the chip drops to below 1.3v and in range of safe current amounts in the processor. Wow, it's like the engineers there know something.

Hang on, lemme get this to you.

TSMC spec at 7nm for 1.3V

Is that big enough for you? It's like you sit on a forum and say "screw all of you" when this is your first 7nm chip. You sit there all high and mighty with what? 14nm? 12? Since when did you make chips better than TSMC and call forward what THEIR chips need?

Again, the difference is the VOLTAGE UNDER LOAD compared to idle. These are not 2700x or older. These are not Intel. Hell at this point you should've gone with a 9900k since Intel can't get their die production fixed after how many years.

Oh, and you're probably having boot issues because of your ram, IF, and something else not being tweaked because "moar voltz moar hertz" at this point. I might have to get myself locked pretty hard from the thread since it keeps popping back up in my feed.

I think it might be time to leave OP to do what he wants with his computer. If he wants to go against established guidelines for whatever reason, and it turns out just fine, then that's obviously great news. If it dies or degrades......well, we tried. It ain't my money, and the risk ain't mine either.

Like a lot of things, it's not always possible to convince someone through reasoned logic when they're bent on believing the opposite.

 

[Toothless]

I think it might be time to leave OP to do what he wants with his computer. If he wants to go against established guidelines for whatever reason, and it turns out just fine, then that's obviously great news. If it dies or degrades......well, we tried. It ain't my money, and the risk ain't mine either.

Like a lot of things, it's not always possible to convince someone through reasoned logic when they're bent on believing the opposite.

Can't fix, well, you know the end of that phrase. Chip burns and RMA? Rejected because ran outside spec. Chip degrades? We'll hear about it in some overly intense thread and all advice ignored. I haven't been this frustrated with someone on the forum in so very long when trying to help.

 

[trickson]

2 or 3 examples is NOT a trend NOR is it proof all I seen is NOT empirical evidence but a bunch of fear mongering. AMD has not officially come and said that there CPU is faulty at 1.4Vcore or 1.35 nor have they changed there max Vcore if any thing it has increased so ...
Some where down the line something went wrong with the MB and CPU more than likely a power spike or surge but you would not know this nor could you test for that and once it happened it took out the CPU again your case is at best weak and filled with holes. You claim this but can NOT prove it definitively NO not at all. Not only that the SYSTEMS are all different not even the same Ryzen 7 3700X CPU. Facts are mounting I call Well you know....
You call me names and say I am stupid and yet I get ban for less?
Any way good luck please keep your fear mongering and BULLSHIT to yourself thank you!
If AMD CPU's cooked at 1.4Vcore under load or NOT then there would be a FUCK TON of RMA's and Returns Not to mention I would NOT have BUILT 3 Ryzen systems!

What kind of BS.... are you trolling? !

 

[Toothless]

If AMD CPU's cooked at 1.4Vcore then there would be a FUCK TON of RMA's and Returns!

This right here if your issue. 1.4v with no load. No current. Nothing going on. Do you not understand how current works? How processor, work, works?

Also, five examples, on the first page of a basic web search. Please show me the research that proves majority that 1.4v is okay under load. We've seen failures already with proof and it's just going to keep growing with more people doing what you do.

It's not fear mongering. It's not bullshit. This isn't Intel 14+nm or Ryzen 2xxx where they can take more voltage. All we've suggested, advised, shown proof, and tried to get you to understand the simple difference of current under load, and current on idle. It's so black and white that it's frustrating. Do you think we're here to make you suffer? I can tell you from testing that many, MANY chips will do 4.4ghz on 1.3v boot and a 0.15v vdroop or better.

One last time mate.

High voltage on idle? All good
High voltage on load? Not good.

 

[trickson]

I think it might be time to leave OP to do what he wants with his computer. If he wants to go against established guidelines for whatever reason, and it turns out just fine, then that's obviously great news. If it dies or degrades......well, we tried. It ain't my money, and the risk ain't mine either.

Like a lot of things, it's not always possible to convince someone through reasoned logic when they're bent on believing the opposite.

Established by whom and on what kind of setups and if all this is so true then would there NOT be a MASSIVE amount of Chips being reported 3-4 posts is hardly a established guide.

This right here if your issue. 1.4v with no load. No current. Nothing going on. Do you not understand how current works? How processor, work, works?

Also, five examples, on the first page of a basic web search. Please show me the research that proves majority that 1.4v is okay under load. We've seen failures already with proof and it's just going to keep growing with more people doing what you do.

It's not fear mongering. It's not bullshit. This isn't Intel 14+nm or Ryzen 2xxx where they can take more voltage. All we've suggested, advised, shown proof, and tried to get you to understand the simple difference of current under load, and current on idle. It's so black and white that it's frustrating. Do you think we're here to make you suffer? I can tell you from testing that many, MANY chips will do 4.4ghz on 1.3v boot and a 0.15v vdroop or better.

One last time mate.

High voltage on idle? All good
High voltage on load? Not good.

Oh see there you go condescending little .... You think YOU know it ALL you can LEAVE NOW! I am trying to be nice and tell you and show you my results. I never asked for YOU to flame me about it nor make me feel like shit about it either You really are a know it all.

 

[Toothless]

Established by whom and on what kind of setups and if all this is so true then would there NOT be a MASSIVE amount of Chips being reported 3-4 posts is hardly a established guide.

Oh, you know, TSMC, who makes the chips, setting the maximum voltage to be pushed through 7nm things, by the engineers who made it.

 

[freeagent]

There is no way I would put any kind of heavy load on my CPU with 1.4v. 1.4v manual vs 1.4 auto or whatever are 2 completely different things lol. Its brutal not suggested.

The patience being shown is astounding..

 

[Toothless]

There is no way I would put any kind of heavy load on my CPU with 1.4v. 1.4v manual vs 1.4 auto or whatever are 2 completely different things lol. Its brutal not suggested.

That's what we've been trying to get across. You can set 1.4v and have a vdroop to 1.3v on load and be perfectly fine. That's exactly what the chip does on stock.

 

[trickson]

That's what we've been trying to get across. You can set 1.4v and have a vdroop to 1.3v on load and be perfectly fine. That's exactly what the chip does on stock.

I do NOT have it set to 1.4 AUTO setting does that and NO it don't drop when under load it does some like 1.35 but nothing bellow this is WHAT THE FUCK I AM SAYING OKAY!
I set the Voltage to 1.35 so I fail to see this issue here. Other than you are right and wrong at the same time being hurtful and condescending is not a way to handle things.
Oh and there is vdrop too so where is it I am going wrong here other than getting 4.4GHz 24/7 and the haters hating on me for that?

 

[Toothless]

I do NOT have it set to 1.4 AUTO setting does that and NO it don't drop when under load it does some like 1.35 but nothing bellow this is WHAT THE FUCK I AM SAYING OKAY!
I set the Voltage to 1.35 so I fail to see this issue here. Other than you are right and wrong at the same time being hurtful and condescending is not a way to handle things.

I'll throw an apology in there if it helps. Do understand how frustrating this is to get one point across.

How about this. Save your profile and try these settings.

Ram to 3200mhz, 1.325v on core, and level 3 LLC.

If it posts, then try a bench. If not, then have it your way.

 

[trickson]

I'll throw an apology in there if it helps. Do understand how frustrating this is to get one point across.

How about this. Save your profile and try these settings.

Ram to 3200mhz, 1.325v on core, and level 3 LLC.

If it posts, then try a bench. If not, then have it your way.

Okay I will give it a try and get back when I do.
My new case came in today and I will be transferring this system into the new one.
Thank you.

 

[freeagent]

That should be ok? I read these cores are meant for 1.3v so you should be able to squeeze a little more out.. I guess that 20% would be 1.325v and there you go..

 

[trickson]

That should be ok? I read these cores are meant for 1.3v so you should be able to squeeze a little more out.. I guess that 20% would be 1.325v and there you go..

Well at present 4.4Ghz VCORE 1.35!!!!!!!! with the RAM at 3466MHz is IMHO nothing to sneeze at. I have never gotten a CPU this high ever and WOW it is fun! And this CPU only cost me $250.00 Bucks!
What a Deal! Amazing CPU for the price Intel could NOT even come close sorry!