Zen 2 (Ryzen 3000) Voltage/Heat Problems Mega Thread [Updated 30/07/19]

[Master Disaster]

Intro

 

It seems like everyone is now buying into Zen 2 (myself included) and everyone (myself included) who has bought in is noticing what seems to be high voltage readings while your machine is sat idle. I've noticed a huge amount of topics being created surrounding this issue so I decided to create this thread where hopefully we can all pool together and share discussion.

 

I've personally done lots of research on this issue and testing on my 3800X and I know others here have too.

 

What exactly is the issue?

 

Zen 2 CPUs seem to be using A LOT of voltage while at stock settings, we're talking 1.4v plus at all times while idling at desktop, I've personally seen mine go over 1.5v and I've heard reports of 1.6v plus from some people. Temperatures also seem to be higher than expected, I can personally say my 3800X is around 10c hotter than my 2700X was in the exact same system, same cooler and actually better heat paste. Some people (myself included) are also experiencing fans in the system randomly increasing and decreasing in speed and finally some users are reporting that their system is powering down after a few hours.

 

 

Have AMD responded to these concerns?

 

They have indeed. Over at this reddit post an AMD employee has made the following detailed post about these concerns...

(PLEASE READ THIS POST VERY CAREFULLY AS IT CONTAINS IMPORTANT INFORMATION)

Quote

Hi, everyone. I've spoken to many of you publicly or privately over the past 48H to better understand why you are seeing idle voltages the community considers to be high. Some of the back-and-forth was covered in this thread, but I wanted to submit my own post to bring more visibility to this topic. We have a final answer for you.

 

Understanding What's Going On

We have determined that many popular monitoring tools are quite aggressive in how they monitor the behavior of a core. Some of them wake every core in the system for 20ms, and do this as often as every 200ms. From the perspective of the processor firmware, this is interpreted as a workload that's asking for sustained performance from the core(s). The firmware is designed to respond to such a pattern by boosting: higher clocks, higher voltages.

 

The Effect of This Pattern

So, if you're sitting there staring at your monitoring tool, the tool is constantly instructing all the cores to wake up and boost. This will keep the clockspeeds high, and the corresponding voltages will be elevated to support that boost. This is a classic case of observer effect: you're expecting the tool to give valid data, but it's actually producing invalid data by virtue of how it's measuring.

 

What about Ryzen Balanced vs. Windows Balanced Plan?

By now, you may know that 3rd Gen Ryzen heralds the return of the Ryzen Balanced power plan (only for 3rd Gen CPUs; everyone else can use the regular ol' Windows plan). This plan specifically enables the 1ms clock selection we've been promoting as a result of CPPC2. This allows the CPU to respond more quickly to workloads, especially bursty workloads, which improves performance for you. In contrast, the default "Balanced" plan that comes with Windows is configured to a 15ms clock selection interval.

 

Some have noticed that switching to the Windows Balanced plan, instead of the Ryzen Balanced Plan, causes idle voltages to settle. This is because the default Balanced Plan, with 15ms intervals, comparatively instructs the processor to ignore 14 of 15 clock requests relative to the AMD plan.

 

So, if the monitoring tool is sitting there hammering the cores with boost requests, the default plan is just going to discard most of them. The core frequency and clock will settle to true idle values now and then. But if you run our performance-enhancing plan, the CPU is going to act on every single boost request interpreted from the monitoring tool. Voltages and clock, therefore, will go up. Observer effect in action!

Okay, Rob. Shhhhh. Just Tell Me How I See Voltages? I Just Wanna Check!

CPU-Z does an excellent job of showing you the current/true idle core voltage without observer effect. In my example image, I've configured a Ryzen 9 3900X with all the same things we would advise the public to use: Windows 10

 

May 2019 Update, the latest BIOS for the Crosshair VIII, and chipset driver 1.07.07 (incl. the AMD power plan). Yes, we're monitoring the behavior of the core, but we can see that idle voltage looks great. The tool is not compelling the firmware to boost when it's not needed.

 

Is There Anything Else I Need To Know?

Yes, actually. The Ryzen CPU depends heavily on a low-power state called cc6 sleep. In this sleep state, core clockspeeds and voltages are basically nil as the core is sleeping and gated. It is not possible to report out the state of the core in this sleep state without waking the core, probing the status, and killing the power savings of cc6. Therefore, MOST tools can only show you the last clock and voltage of the core before the core went to cc6. So if you were at full 4.5GHz+ boost @ 1.48V, then the core went to sleep, many tools might show the core(s) stuck at that value. The tool just doesn't know any better.

However, the latest version of AMD Ryzen Master can uniquely show you clocks and voltages in a cc6 state. No other tool can do it. Neat piece of info for the people looking to understand how their core behaves!

tl;dr: Observer effect bad. You can't always trust your tools. CPU-Z gives you the right idle voltage. We'll look at the rest. Thank you everyone for your reports and insight, which helped us get to the bottom of this once and for all.

//EDIT: To ensure you're following my instructions correctly:

1. Do not have two different monitoring apps running to compare them, e.g. Ryzen Master and CPU-Z. Or CPU-Z and HWINFO. I see many folks trying to run two apps at the same time, so they can compare behavior. This can cause a race condition, which will affect your results.

2. Just run CPU-Z at the desktop, by itself, with no other monitoring apps going.

3. Don't forget background apps like Corsair iCue, NZXT CAM, or software that came with your mobo are also monitoring tools.

4. Make sure all BIOS voltage settings are set to NORMAL or AUTO. Only enable your XMP profile for the purposes of this test.

5. Make sure you have chipset driver 1.07.07 (from amd.com), Windows 10 v1903, and the latest BIOS for your motherboard.

6. Do not worry if your processor is not exactly matching mine with voltage. All we're looking for is the CPU to go to < 1.0V when you're staring at CPU-Z doing nothing. This indicates idle is workig correctly.

7. If you are 100% convinced that you've followed my steps correctly and you're still seeing 1.38V+ idle voltages, PLEASE FILL OUT THIS FORM (it's anonymous!).

//EDIT @ 07/12/2019, 00:14 UTC:

I'm specifically looking for reports where the voltage is stuck at a particular value, or a small range of values, around 1.4V--no matter how long you sit there and watch it. It is perfectly okay if your CPU is periodically using 1.4-1.5V to achieve boost frequencies, and you should see dips into sub-1.0V as the CPU goes into idle. These dips may be brief, and that's okay. Load voltages of around 1.2-1.3V are perfectly okay also. This is the processor working as expected. Ryzen is a highly dynamic system, with up to 1000 voltage and clockspeed changes every second. You will see a lot of bouncing around as you work with your system.

I anticipate that many people are now trying Ryzen processors for the first time (because they're awesome), and may not understand what to expect versus whatever CPU they had previously. You want to know if what you're seeing is "normal," but may not know what "normal" looks like. I get it! I want to assure you that the CPU needs voltages to boost, and voltages of 1.2-1.5V are perfectly ordinary for Ryzen under load conditions (games, apps, whatever). Even at the desktop, Windows background tasks need love too! You'll see the CPU reach boost clocks and voltages, too. But if your voltage is well and truly stuck, that's what I'm trying to troubleshoot.

 

EDIT 7/13/19 @ 18:28 UTC If your BIOS has the option to set CPU voltage to AUTO or NORMAL, please try setting it to normal. Please also make sure you've installed chipset driver 1.07.07 from amd.com. I have received reports from several people that this resolved their issue. We continue to diagnose the reports, though, and appreciate the data coming in from the community!

 

EDIT 7/18/19 As a temporary workaround, you can use the standard Windows Balanced plan. Edit this plan to use 85% minimum processor state, 100% maximum processor state. (Example). This will chill things out as we continue to work this issue. Your 1T and nT scores shouldn't change at all (+/- the usual run-to-run variance). This will preserve boost, retain cc6 core sleeping, preserve idle downclocking/downvolting, but make the CPU more relaxed about boosting under light loads.

Please note that it is totally normal for your Ryzen to use voltages in a range of 0.200V - 1.500V -- this is the factory operating range of the CPU. It is also totally normal for the temperature to cycle through 10°C swings as boost comes on and off. You will always see these characteristics, as they're intended, so do not be surprised to see such values.

Please do not undervolt the chip or set a maximum processor state of 99%. These are ineffective and/or detrimental changes.

We appreciate the reports everyone has provided, and they are helpful. I will make an all-new post when I have a more comprehensive update to share. Thanks for your patience. ♥

EDIT 7/22/19 Hope to have an update for everyone, soon. I will make a new thread for it. Thank you again for your patience. I've received kind messages of support over the past week, and I really appreciate it. I know people are eager to hear more. Soon.

 

What happens now then?

 

The first thing you do is try Michael's recommendations from above. That is as follows

1. Only use Ryzen Master to monitor voltages, close all other monitoring apps
2. Make sure the Ryzen Balanced profile is loaded in Control Panel > Power
3. Make sure that voltage is set to either Auto or Normal in BIOS, if Normal is present use that otherwise use Auto
4. Do not try and undervolt the CPU manually
5. Make sure Corsair iCue, NZXT Cam and GeForce Experience are not running, these apps are known to cause Zen 2 to never idle

After following these recommendations close all background applications you possibly can and then run Ryzen Master and leave your computer alone for 5 minutes, do not touch the mouse or run anything. Watch the Core Voltage reading and check if it is dropping below 1.4v at all, it doesn't matter if it only drops for a split second, as long as it drops your CPU is operating normally.

 

I followed these steps and my Voltage doesn't drop below 1.4v?

 

Don't panic, these steps did not fix my CPU either.

 

First thing, please head over here and fill out this survey for AMD.

 

Many people are experiencing these issues and many are reporting "fixes", what I will do is list them here. Remember that each oone is a standalone fix, try the steps outlined exactly then follow the instructions below to check if it worked for you

 

Quote

After following these recommendations close all background applications you possibly can and then run Ryzen Master and leave your computer alone for 5 minutes, do not touch the mouse or run anything. Watch the Core Voltage reading and check if it is dropping below 1.4v at all, it doesn't matter if it only drops for a split second, as long as it drops your CPU is operating normally.

 

Fix 1 (This one worked for me and multiple other people here, try this first)

1. Open Control Panel > Power Options
2. Select the High Performance profile
3. Click Change Plan Settings
4. Click Change Advanced Power Settings
5. Scroll to Processor Power Management and click the Plus Sign
6. Set Minimum State to 1% and Maximum State to 100%
7. Click OK and close the window
8. Repeat the testing and check your Voltage is dropping

Fix 2

1. Open Control Panel > Power Options
2. Select the AMD High Performance profile
3. Repeat the testing and check your Voltage is dropping

Fix 3

1. Disable Performance Boost Override in BIOS

Fix 4 (Not recommended by AMD)

1. Set a voltage offset of minus 0.1v in BIOS (-0.1v)

Will these issues be fixed in the future?

 

Here's the funny thing, we don't actually know if this is an issue or not? Its very unlikely AMD intended for Zen 2 to suck up 1.4v at idle but the platform is so new that its entirely possible it is behaving as intended and we're all just worrying over nothing.

 

The second point is about managing expectations. Every person is different and what they expect from a platform is also different, personally I expect my CPU to idle and not run at 40c while literally doing nothing. I don't think that's an unreasonable expectation nor is it one i expect to be unpopular but remember, just because we think its not operating normally doesn't mean that its not operating normally.

 

Now to answer the question, if this is indeed and fault then I would expect AMD to fix it in firmware, that would mean an AGESA update which would be delivered to us via a BIOS update from our motherboard manufacturers. Zen 2's firmware is still very much in BETA anyway so more AGESA updates will be coming no matter what. Keep an eye on your boards support website for BIOS updates, you should be installing anything that comes out but ITS VERY IMPORTANT YOU INSTALL ANY BIOS THAT HAS AN AGESA UPDATE INCLUDED.

 

If and when any AGESA update is released I will update this thread with details.

 

Hopefully this thread can help out some people, if you have a concer then please post it, as well as myself there are lots of users here running Zen 2 that will be happy to help.

 

Thanks for reading.

 

Edit 30/07/19

 

Lots of fresh information and a new driver release from AMD.

 

Before updating please do the following...

1. Open Control Panel > Power Options
2. Click Advanced Settings
3. Select any default Windows profile
4. For any Ryzen profile click Change Profile Settings
5. Click Remove Profile
6. Repeat steps 4 & 5 for all Ryzen profiles

After you've removed all the Ryzen profiles install the updated driver from here.

 

There's also been an update to Ryzen Master and AMD now recommend you use only that to monitor temps and voltage.

 

Get it from here

 

Finally a new AGESA update has been sent to board manufacturers and is currently undergoing testing so please keep an eye out for a new bios release in the next few days.

[Zando_]

*follows because I'm still debating getting an ITX board for my i5 8400 or a Zen 2 chip for my B450 ITX board*

 

Also please fix the quote text, it's set to black so those on dark mode like me can't read a thing, lol. 

[Master Disaster]

I forgot to say, anybody with any fixes I missed please post em and I'll do my best to add em to the OP.

[Master Disaster]

1 minute ago, Zando Bob said:

*follows because I'm still debating getting an ITX board for my i5 8400 or a Zen 2 chip for my B450 ITX board*

 

Also please fix the quote text, it's set to black so those on dark mode like me can't read a thing, lol. 

I don't remember how?

[Zando_]

Just now, Master Disaster said:

I don't remember how?

I think you can select the text and hit the Tx button next to the emoji one, that clears the formatting so it should automatically go black on light theme and white on dark theme. 

[Master Disaster]

Just now, Zando Bob said:

I think you can select the text and hit the Tx button next to the emoji one, that clears the formatting so it should automatically go black on light theme and white on dark theme. 

thanks

[Zando_]

5 minutes ago, Master Disaster said:

thanks

Awesome! Can read it now!

[paulmohr]

Well I for one am not having any issues with mine. 3600x on a b450 tomahawk and it seem to work quite well. Nothing that far off from what my 1500x was honestly, except it obviously performs much better. I tested both with a cheap deepcool air cooler and then I upgraded to a 240mm H100i pro AIO cooler. The water cooler works a little better, but nothing amazing to be honest. At idle it runs around 35 to 40c and under a load it hovers around 70c. And my ambient temps are high, I live in an attick in Michigan, lately its been between 90-96 F in here. With an all core overclock it reaches a little more than 75c under a load (cinabench). Even when I overclcocked the shit out of it and was pulling nearly 120 watts it only hit like 93c under load.

 

Something to keep in mind, like I said in another post is how we are monitoring these parameters. We are using the built in sensors which I would assume are the cheapest sensors they could find and not really designed to take the readings we are trying to look at. And we are using software that is known to not be all that accurate either. We can get estimations from this method but its far from accurate or reliable measurements for testing. In order to do this properly you would need equipment to measure voltage, current and proper temps at the correct places that is actually made this purpose. Which some people have and have used, and have the training to use such equipment. Especially when we are talking about heat. We are seeing swings in the 30c range in a split second. I have never seen anything actually do this. It takes time to generate heat and dissipate it. I think we are seeing false or inaccurate readings because of our method. I could be wrong though, I am not an electrical engineer. My background for testing equipment was in the automotive and fighter jet fields.

 

I watch videos and read articles where reviewers and technicians have tested these products with the proper equipment and they all for the most part talk about how well they work and how efficient they are. Then I come to forums and see people talking about how power hungry and hot they run. The two seem to contradict each other and don't make much sense honestly. All I know is what I see. And what I see is a product that runs at higher clock speeds, uses less wattage, runs at basically the same temps and somehow has a lower rated tpd rating and draws less current than my 1500x did. Maybe I just got really lucky, I don't know.

[Ravendarat]

My temps were always fine on my 3600 (custom loop so it’s got help) but I checked my voltage with ryzen master today and saw it idled at 1.48 which is obviously on the high side, I did fix one and it fixed it right up to frequently dropping down as low as .96x while idling which according the reddit article is what I was looking for

[caldrin]

3800x here on an Asus x570 TUF. 

 

If the CPU ever gets to an idle state temps are usually in the mid 40s.. using an Kraken X62 AIO for cooling, but most of the time the CPU is not idle even when the computer is doing very little so temps usually sit around the 55 to 60 mark and every now and then jumps up to 70.. Under load and gaming it can be anything from mid 60s to mid 70s but have seen it spike above 80 once or twice.. Cant recall the voltage hitting 1.5 but will keep an eye on it.

Running the AMD performance power plan and have not tried any of the above fixes. Latest updates for Windows and bios fully up to date.. I guess they just run hot as everyone seems to be having a similar story. 

[Dubba]

3700X here on an Asus x470-F

 

Monitoring is indeed difficult right now. I swapped to the newest Afterburner Beta and disabled almost all the CPU monitoring options and left only overall usage and clock visible. I think I can trust the usage number but clocks are really different from what either CPU-Z or Ryzen Master tells me. I want to keep Afterburner open for GPU stuff, any idea if it conflicts with other CPU monitoring software if I disable those options?

At one point I had Afterburner and Ryzen Master open at the same time and while RMaster showed clock speeds of about 200MHz in idle, Afterburner never showed anything below 2GHz. Idle voltages dropped to around 0.9V but only for a second at a time, according to RMaster.

 

I did some benchmarking with only Afterburner monitoring all I wanted from GPU and only CPU usage & overall clock. Games, TimeSpy and Cinebench single threaded showed clocks of 4250-4375MHz and hard all core loads gave me around 4050MHz, seems legit. But I'm gonna do a quick double check with only Ryzen Master monitoring.

 

In idle and in gaming loads I see voltages of 1.4-1.5V, in all core loads it's pretty stable between 1.39-1.41V

Only setting I changed in BIOS was to enable PBO. Windows is in Ryzen Balanced.

 

Temps are obviously pretty hilarious when compared to the expectation of this new power efficient 7nm process

Idles of 50-60°C, gaming at 70°C+, max load 84°C, under NH-D15 and an ambient temp of 31°C. I mean, it doesn't look totally out of place, only that the voltages are too high.

[MrMoonboots]

Hey guys, AMD released a new chipset driver hoping to address some issues with their ryzen balanced power plan
https://community.amd.com/community/gaming/blog/2019/07/30/community-update-5-let-s-talk-clocks-voltages-and-destiny-2

[MrMoonboots]

So, now after the chipset driver install the balanced power plan uses 1.45-1.5 on average and I see no drops below 1V
So much for that fix. Anyone else having a similar result, or?

Just went back to Windows High Performance because it's still the only working solution.

[Master Disaster]

47 minutes ago, MrMoonboots said:

Hey guys, AMD released a new chipset driver hoping to address some issues with their ryzen balanced power plan
https://community.amd.com/community/gaming/blog/2019/07/30/community-update-5-let-s-talk-clocks-voltages-and-destiny-2

I'll test it tomorrow, I'm away from my PC right now but thanks for the information.

 

Did you delete the old AMD profiles before updating the driver?

[Gohardgrandpa]

Is destiny 2 even able to launch now? I’d like to play it but unfortunately am unable to after upgrading to the 3600 

[ch3w2oy]

I didn't read everything but my voltages are definitely higher (~1.46v) on my 3700X/3800X than my 2600/X.. But my liquid temperature is actually pretty similar or better..

 

Not sure if that matters but maybe Zen 2 needs more voltage? 

[Master Disaster]

5 minutes ago, Gohardgrandpa said:

Is destiny 2 even able to launch now? I’d like to play it but unfortunately am unable to after upgrading to the 3600 

New driver update contains a temporary workaround to run D2, full fix will be included in the next AGESA update.

[BroliviaWilde]

i'm hesitant about updating chipset driver.  i've got an x470 taichi and a 3900x and just got a couple of issues fixed via rolling back the bios from 3.5 to 3.4.   When I first got the 3900x, my mobo was on 3.4, but the IF would not match RAM speeds, but boost clocks and idel clocke along with idle votages looked just fine.  I even ran ryzen master, cpu and hwinfo, GFE all at same time and I could see voltages drop to about 1, even to .9 sometimes.   updated to bios 3.5 which fixed the IF issue, but then  boost clocks are lowered, idle clocks are never below 3 and voltages stay at 1.4 or higher at all times.  rolled it back to 3.4, now it all works like ti is supposed to:  boost clocks, idle clock, idle voltages, IF, it is all working right, thinking that i'm not going to update to any new bios or chipset driver again unless somethign really does break.

[Master Disaster]

Update 30/07/19

  

Lots of fresh information and a new driver release from AMD.

 

Before updating please do the following...

1. Open Control Panel > Power Options
2. Click Advanced Settings
3. Select any default Windows profile
4. For any Ryzen profile click Change Profile Settings
5. Click Remove Profile 
6. Repeat steps 4 & 5 for all Ryzen profiles

After you've removed all the Ryzen profiles install the updated driver from here.

  

There's also been an update to Ryzen Master and AMD now recommend you use only that to monitor temps and voltage.

  

Get it from here

 

Finally a new AGESA update has been sent to board manufacturers and is currently undergoing testing so please keep an eye out for a new bios release in the next few days. 

[AvogadrosDog]

This is great to hear. My 3700X has also been conforming to the general trend with idle voltages of 1.4v and upwards. But under load on cinebench I get voltages of about 1.187v. Clocks wise I get up to 4.375 GHz with loadings of 3.95 to 4.05 GHz when running cinebench. My only true worry is temps as in my previous builds I could run cinebench with max temps of 60-65c whilst now with my 3700X I’m looking at 71-75c. I’m worried to even try heavier stress tests like Prime95 or AIDA64. 

 

I’m confident that AMD will sort this out or clarify or both in due time. For now I don’t mind attempting different configurations or just plain waiting. Last thing I want to do is burn my chip. - slightly glad to see other people having shut down issues as I thought something odd had happened to me but it seems to be a thing haha. Hope no ones chips got fried though.

[MadmanRB]

Yeah this is very simular to when zen first launched, growing pains to be sure

But hey all new platforms have this so whoop de doo

[Ravendarat]

Yay!! I Destiny 2 worked lol

 

[lansing]

Okay new update seem to have some improvement.

 

At idle, this is the first time I'm seeing below 40C temp on my 3900x, before it always jump back to 50C when it reaches 40. The cores also seem to stay sleeping instead of waking up every other second. Core voltage remains a constant 1.1x v on idle.

 

At 100% load, core voltage stays at 1.28v and temperature never get pass 87.5C

[Dubba]

Just updated chipset driver and Ryzen Master. Desktop temperatures went from 50°C to 40-45°C while idling and scrolling the forums with a dozen tabs open.

While writing this the voltage hovers around 0.9V with some occasional small spikes to 1.1v.

When in an empty desktop cores seem to stay in sleep and voltage stays at 0.7-0.9v.

 

Looks better to me now. But.

I lost 1% in cinebench R20 single- and multithreaded scores!!! Not yet sure if this is because of random boot by boot variance.

Single core boosted to only 4.32GHz and didn't spike to +1.5v like yesterday.

All-core stayed pretty much the same at 4.06GHz. and voltage increased by 0.1v from previous testing to a stable 1.41v and temperatures rose to 84°C again.

 

They did mention that they changed the voltage metering methodology, but I want my 1% of score back

 

[paulmohr]

My cinabench scores vary by more than 1 percent from run to run or depending on the day I run it lol. I wouldn't worry too much about it.