5900X - SVI2 TFN hitting 1.53v with VID 1.55v - AMD support says it's normal?

D

Deleted member 345482

Guest
Please see this thread instead of reading this one. Information is sourced and clearly presented.

Hello,

This is my first post so hope I'm doing it correctly. I apologise if I'm not. I've only had my PC for a about two months and love the performance of the 5900X after coming from a 6700K, but find the voltage rather high on occasion (when it's over 1.500v). I'm using an Asus B550-E with the latest BIOS as of writing with stock settings (Auto vcore, LLC, PBO set to disabled - also tried auto - with RAM set to 3600 CL18, but the same thing happens on 3200, Asus 'enhancements' disabled). I phoned AMD support and they put me through to 2nd level support - I emailed them a couple of screenshots and explained that when I say vcore, I mean the value HWINFO reports as SVI2 TFN. AMD support said it's totally normal to see VID of 1.55v and actual delivered voltage up to 1.55 for brief periods on Zen 3, but the AMD presentation slide for the 5000 series shows "0.200 - 1.500v" is normal and it seems AMD employees on reddit have said 1.500v is the max. For a user coming from a decade of Intel back to AMD (my last AMD CPU was a Turion X2), this is a confusing mess.

From what I can tell (using HWINFO to log data every 5,000ms and checking values), the boosts over 1.500v are spikes that seem to last for one polling cycle. There might be five of these spikes in a five hour period (playing Civilisation VI), or a few spikes in a 45 minute period (Chrome, general desktop stuff).

Would you be concerned, or is this normal behaviour?
 
Last edited by a moderator:
D

Deleted member 345482

Guest
yup. you can dial it down a bit with an offset to keep it simple or as others prefer, make it complicated with curve optimizer.

Thank you for replying. I tried PBO and found it made little difference to real world performance. Stock CBR23 is 21,671 MC and 1,606 SC. The Curve part is confusing - I understand that one step out of the 30 represents between 3mV and 5mV but why not just put a sensible Curve option rather than '30 steps'. Either way, with just a negative offset of 0.0250v, SVI2 TFN still goes over 1.500v. It's a hungry chip. Since you mentioned it, it seems the Curve Optimiser is most popular, but is there anything particularly wrong with setting a negative offset instead?

This is actually normal behavior. These chips will pull slightly more than 1.5v when applying boost clocks.

Thank you also for replying. The attached image is the part that confused me. There are three source for the information - a couple of AMD employees on reddit answering questions and stating 1.500v is the limit for Zen 3, but 1.55v is OK for Zen 2 (I can't find the exact post). Then there's the attached official AMD slide with also states 1.500v. Then there's AMD support which contradicts AMD employees and AMD's own marketing/promotional material.

AMD-Ryzen-5000-Desktop-CPUs_Temps_Overclock_Power-Limit_Profiles_2.jpg

I've also attached another image which shows 45 minutes of data and the boosts over 1.500v - they are minor, but that was during Civilisation VI using ~10 threads. Polled at 2,000ms straight to M.2 NVMe (technically far less load on the CPU, according to HWINFO's developer).
45m-CivVI-SVI2TFN.png

I'm not sure whether full size images are preferred here, or if thumbnails are. It seems full size images are but I don't want to break the rules with my second post.
 

learners permit

[H]ard|Gawd
Joined
Jun 15, 2005
Messages
1,223
Minor voltage overshoot won’t affect your cpus longevity. The 10ms or so at overshoot is there to ensure transient loading stability. Some cpus need it some do not hence the curve optimizer option in your bios that allows you dial up or down by approximately 120mv on the average. My 5900 isn’t a great sample so I run it at -12 all cores and get short bursts of 4950MHz during gaming sessions and 4550 sustained loads. Plenty fast and temps max at 84C.
 

pendragon1

Extremely [H]
Joined
Oct 7, 2000
Messages
43,754
Since you mentioned it, it seems the Curve Optimiser is most popular, but is there anything particularly wrong with setting a negative offset instead?
neg offset isnt as fine tuned, thats all. some will say "youre not doing it right" if youre not using it but whatever, its easy.
 
D

Deleted member 345482

Guest
Minor voltage overshoot won’t affect your cpus longevity. The 10ms or so at overshoot is there to ensure transient loading stability. Some cpus need it some do not hence the curve optimizer option in your bios that allows you dial up or down by approximately 120mv on the average. My 5900 isn’t a great sample so I run it at -12 all cores and get short bursts of 4950MHz during gaming sessions and 4550 sustained loads. Plenty fast and temps max at 84C.
My CPU must be a terrible sample as the 'worst' core of CCD0 (core 5) throws errors at anything more than - 7. If the voltage is considered normal, I'll stop worrying about it and put the confusion down to AMD's marketing vs technical.
 
D

Deleted member 345482

Guest
After more testing, Core 0 (my best core, according to Ryzen Master and HWINFO - boosts to 4.95 GHz at stock and sustains 4.9 GHz) fails using CoreCycler at iteration 18 with just one more bump in negative offset (taking it to 5 steps) - ~0.03500v.
 

Dreamerbydesign

Supreme [H]ardness
Joined
Feb 3, 2008
Messages
5,798
My CPU must be a terrible sample as the 'worst' core of CCD0 (core 5) throws errors at anything more than - 7. If the voltage is considered normal, I'll stop worrying about it and put the confusion down to AMD's marketing vs technical.
You have the answer right there. Also yes these are power hungry chips. They also do run a bit warmer than most of us were used to.
 
D

Deleted member 345482

Guest
You have the answer right there. Also yes these are power hungry chips. They also do run a bit warmer than most of us were used to.
My 6700K idled at about 40°C with a 280mm radiator (CLC) and up to 90°C under load - that was with the terrible (and now very old) Corsair Obsidian 750D. My 5900X system idles between 35-42°C and under load (x265 encoding, two jobs at the same time = 100% CPU use) stays at ~65-70°C. The difference is gaming, where only a few cores at used and therefore more voltage - that's where I see temps of ~71°C even on games like CSGO. Still, the heat of these chips is fine (IMO) - it's only AMD contradicting AMD that concerns me as I didn't know which answer is correct.
 

Dreamerbydesign

Supreme [H]ardness
Joined
Feb 3, 2008
Messages
5,798
My 6700K idled at about 40°C with a 280mm radiator (CLC) and up to 90°C under load - that was with the terrible (and now very old) Corsair Obsidian 750D. My 5900X system idles between 35-42°C and under load (x265 encoding, two jobs at the same time = 100% CPU use) stays at ~65-70°C. The difference is gaming, where only a few cores at used and therefore more voltage - that's where I see temps of ~71°C even on games like CSGO. Still, the heat of these chips is fine (IMO) - it's only AMD contradicting AMD that concerns me as I didn't know which answer is correct.
I e had a 5900x also for months now. I saw temps peak around 80c on a 360 Corsair aio. Idle, around 32-38 generally. I’m used to the temps now. I think many of us freaked out when we first got our hands on these. But in the end the temp is fine.
 
D

Deleted member 345482

Guest
A new development - overclocker 'The Stilt', who I imagine is also registered here - has stated and confirmed that the maximum voltage the 5900X (all 5000 series processors) is hard capped at 1.50000V when under true stock conditions. Seeing a VID higher than 1.50000V is not normal.

This means the motherboard (at default settings) is still doing something to adjust the CPU behaviour. We've gone through telemetry, common overclocking options, etc.

He said he's going to look into the BIOS to try and work out what Asus is doing that stops the CPU from operating at AMD-defined specifications.

Just thought many here would find this information useful and interesting.
 

pendragon1

Extremely [H]
Joined
Oct 7, 2000
Messages
43,754
Following on from this thread, I've continued to do research and spoken with the AMD Technical Marketing guy who made the Zen 3 voltage slide, as well as overclocker The Stilt. Both have opposing views of what is acceptable, with the latter basing it on AMD's baked-in limits...

A little bit of context partly copied from my thread at OCN. It's a simple question, but after more than 5 weeks of searching for an authoritative answer, I'm lost. In theory, AMD should be that authoritative answer but AMD's hardware contradicts its employees (more below).

The Question
At stock power limits with PBO and any vendor-specific motherboard 'enhancements' disabled, as well as automatic voltage and LLC, and with DOCP/XMP enabled or disabled, what is the maximum amount of voltage Zen 3 can request and will be delivered?

The Reason For Asking
The reason for asking this simple question is after noticing HWiNFO reporting up to 1.531V being delivered (VDDCR_CPU/SVI2 TFN) to the CPU, and 1.550V being requested (VID).

Now on to the answers. In order to make it easier to read, I will insert thumbnails instead of full-sized images. If people find this difficult to see, please let me know and I'll edit the post.

Answer #1: AMD promotional slide - the following is taken from the thumbnail side

Answer: 1.500V

Answer: 1.500V
Source:
View attachment 419976

Answer #2: AMD Support
- the following is taken from an AMD support email

Answer: 1.531V - 1.550V
Source #1:
View attachment 419977

Source #2: Phone call where I mention 1.550V, which unfortunately I can't provide a copy of.

Answer #3: Professional overclocker 'The Stilt' - the following is taken from a discussion on Reddit and OCN with overclocker The Stilt

Answer #1: 1.50000V
Note: After speaking privately with The Stilt and probing my CPU (following his instructions), he confirmed the CPU is hard-limited to 1.50000V. Therefore, it's my Asus B550-E that is doing something, although after The Stilt spending a day working on different BIOSes and boards, he couldn't work out what.
Source:
View attachment 419979


Answer #2: 1.50000V (delivered voltage)
Source:
View attachment 419981

And finally, confirmation of what the CPU reports
View attachment 419982

Answer #4: AMD Technical Marketing 'AMD Robert' - the man who created the slide in Answer #1 - the following is taken from a discussion on Reddit with verified AMD employee Robert, who works in Technical Marketing and answers technical on the AMD sub-reddit.

Answer: 1.550V (VID and delivered, i.e. VDDCR_CPU)
Source:
View attachment 419983

Conclusion
Assuming everything is stock (excl. XMP/DOCP);
  1. based on AMD's slide the maximum VID is 1.5000V
  2. based on the person who made the slide the maximum VID is 1.5500V
  3. based on AMD Support the maximum 'voltage' is 1.5500V
  4. based on The Stilt, a world-famous professional overclocker, the maximum VID is 1.5000V
  5. Based on my specific CPU, the maximum VID is 1.50000V
What are your thoughts? AMD's hardware contradicts AMD's statements...

The Stilt confirmed the exact CPU limit for voltage requests on my CPU. That limit, according to the CPU, is 1.50000V. That is, according to The Stilt, the case with at least the 5800X, 5900X, & 5950X. I realise this only applies to VID, but a higher VID often leads to a higher Core SVI2 TFN.
didnt need to start a new thread. its normal as per the quotes you provided.

1639015649006.png
 
D

Deleted member 345482

Guest
didnt need to start a new thread. its normal as per the quotes you provided.

View attachment 419991
Oh. Sorry. To me it seemed creating a new thread was the right thing to do as information has changed since I created this one. I've now got hardware confirmation of the factory limit of my CPU (1.50000V) which contradicts AMD's statements. The motherboard is therefore doing something (at stock) to the CPU. The Stilt spent quite some time going through multiple BIOSes on two boards to try and find the cause. Asus, in this instance, is not honouring AMD's factory voltage limit. To me, that's significant and so is his third sentence (third line down)... unless The Stilt is incorrect but I trust he knows what he's doing. stilt-4.png
 
Last edited by a moderator:

learners permit

[H]ard|Gawd
Joined
Jun 15, 2005
Messages
1,223
So I get that this is unsettling to you in some manner. Why not set your own happy limit in bios and settle it yourself? A little testing is all it would take to do so since you already have command of the necessary monitoring software and control of your own bios.
 
D

Deleted member 345482

Guest
So I get that this is unsettling to you in some manner. Why not set your own happy limit in bios and settle it yourself? A little testing is all it would take to do so since you already have command of the necessary monitoring software and control of your own bios.
It should be unsettling to any user with a VID over 1.5000V.
 
Top