2700X high idle temps?

LigTasm

Supreme [H]ardness
Joined
Jul 29, 2011
Messages
6,610
Wondering if anyone else with an X370 board is seeing this? My 2700X at stock with XMP off idles at 50C and constantly surges to 65C and back causing the CPU fans to constantly be ramping up and down.

Mount is good, and under load the temp actually goes down. I notice idle voltage is like 1.550V while once it loads and goes to turbo (I usually see 4100-4250mhz) the voltage drops DOWN to ~1.280 and the temps actually go to the mid-40's then.

I tried updating the BIOS to the one after the Zen+ compatability (4008 to 4011 on X370-i Strix), no change. I tried using normal balanced mode instead of Ryzen balanced mode as well.

The only thing that has fixed it so far is manually setting the voltage lower but then it does not boost at all, it just sits at 3700mhz.
 
try messing with the windows power plan, set the minimum to 5% and see what it does..

also not sure if this is the case but from the few threads i've checked else where some one noticed the high idle voltage eventually went away over time so it's possible there's something in the bios that eventually learns what the voltage should be at idle, but i'm not 100% sure. hard to tell if they actually changed something and just didn't bother saying what they changed.
 
try messing with the windows power plan, set the minimum to 5% and see what it does..

also not sure if this is the case but from the few threads i've checked else where some one noticed the high idle voltage eventually went away over time so it's possible there's something in the bios that eventually learns what the voltage should be at idle, but i'm not 100% sure. hard to tell if they actually changed something and just didn't bother saying what they changed.

I did mess around with the power plans, but the high voltage persists even in BIOS. Its strange that it actually drops under load. If I have time I may put it in one of my other AM4 boards and see if it behaves the same, because listening to the fan surge up and down all day is extremely annoying.
 
I had this problem with my 1800X, solution in my situation was to drop in a 1700 instead. Worked a lot better. Now I have a 2700X (assembling tomorrow). 2700X improved a lot on voltages, so I have to wonder about your board. Fan speed is the really annoying part about this, I can turn a blind eye to high temps, but not to fans going crazy. If you can't get a flat fan speed in your UEFI, then buy a fan with whatever fixed RPM you can stand and put that in there. It'll either throttle back or be fine either way.

I also think the 1800X and 2700X are likely to simply be pushed to the absolute max, they really aren't great chips to be buying if one is like me and you and meticulously monitoring voltages. My opinion is to use it, enjoy it, if it remains after a year of UEFI updates, buy yourself a 65W TDP 7nm Ryzen3. That's what I'm going to do if mine doesn't behave.

Also something relatively useful I found out in my comparisons, the 1700 ran about -10C across the board from the 1800X in the same system. So instead of idling at 40C (wouldn't be uncommon), idling at 50C seems about to be expected. You have to hit 60C for AMD's throttling algorithm to kick in a way that affects your XFR2 curve with temps, but just use it.

In sum yes the 65W TDP chips are almost always the best choice. I just can't resist trying to make the dumb choices like the 8700K and 2700X work though.
 
I had this problem with my 1800X, solution in my situation was to drop in a 1700 instead. Worked a lot better. Now I have a 2700X (assembling tomorrow). 2700X improved a lot on voltages, so I have to wonder about your board. Fan speed is the really annoying part about this, I can turn a blind eye to high temps, but not to fans going crazy. If you can't get a flat fan speed in your UEFI, then buy a fan with whatever fixed RPM you can stand and put that in there. It'll either throttle back or be fine either way.

I also think the 1800X and 2700X are likely to simply be pushed to the absolute max, they really aren't great chips to be buying if one is like me and you and meticulously monitoring voltages. My opinion is to use it, enjoy it, if it remains after a year of UEFI updates, buy yourself a 65W TDP 7nm Ryzen3. That's what I'm going to do if mine doesn't behave.

Also something relatively useful I found out in my comparisons, the 1700 ran about -10C across the board from the 1800X in the same system. So instead of idling at 40C (wouldn't be uncommon), idling at 50C seems about to be expected. You have to hit 60C for AMD's throttling algorithm to kick in a way that affects your XFR2 curve with temps, but just use it.

In sum yes the 65W TDP chips are almost always the best choice. I just can't resist trying to make the dumb choices like the 8700K and 2700X work though.

Well, my goal was since its a mini ITX rig I wanted the maximum boost without fiddly overclocking and trying to deal with voltages constantly. It took a month of tweaking to get my 1800X stable at a speed I liked and this 2700X beats it out of the box. I would have just bought the X470 Strix board but its out of stock everywhere so I stuck with the X370.

Anyways, this pissed me off enough that I ordered blocks and radiators, just going to water cool it and be done with it. The Vega card makes some noise in games but I wear headphones then and don't notice it. The CPU is currently under a Noctua NH-U12S with two fractal venturis on it, very quiet fans. But the constant ramping is really grating on my nerves, I can't seem to get it to manually stay at one speed unless I just use an adapter off of a molex plug but I don't want all the extra wire clutter. I played with the fan hysteresis and it doesn't seem to do squat in Asus fan control, I really like ASRock and Gigabyte much better for fan options.

So now it'll be under water by the end of the week and we'll see what happens then.
 
Deja vu reading your posts, reminds me with my 1800X back in March 2017. Some (but not all) boards use the +20C offset sensor for the fan, so you have very little wiggle room on temps before it spins the fan up if you have one of those.

I have the Asus x470 ITX arriving tomorrow myself. In my opinion you have a fan issue and nothing else. I ultimately regretted swapping my 1800X for a 1700, it was dramatically improved, but the fan was still changing speed more than I wanted. I think it would've been fine with the Noctua NF-B9 PWM on the 1800X, that I ended up putting on the 1700 anyway. I disable fan controls so it's running fullspeed (1600RPM).
It works, but you're going to get better performance out of your watercooling solution in the end so that certainly works too.

With games or heavy load, you're gonna have noise, the GPU and radiator fans will spin up. That situation is tough to avoid, I wouldn't ever let that bother me. My only requirement is that the fans aren't going up and down on the desktop opening Firefox. Totally agree with you, it drives someone nuts. I'm also onboard with the 2700X because it's really nice to get maximum single core performance, 4.35GHz without having to fiddle with it.
 
What you're describing (assuming you do indeed have no mounting/hardware issues or lack of sufficient cooling) is probably one of three things:

- Your vdroop compensation is at insane levels (check your LLC, try it at 5*)
- You have a manufacturer custom-specced PBO profile that overvolts your chip (lower it down if activated; if it's not, take it out of Auto and start by setting it to 2)
- A combination of both

* This is only for Asus owners, MSI's LLC levels are backwards.

FIY, all Ryzens behave in this way, albeit of course not to extremes; they have a higher voltage at idle and a noticeably lower one during resource intensive tasks; this is due to Amperage increase and actual chipset instructions. In your case, it's the scale of this, not the effect per se.
In the future, avoid Strix motherboards. They make them as cheap as they can get away with, but charge you for the logo nonetheless. This isn't 1998, being a victim of hype is the consumer's fault and the consumer's fault alone.
 
Last edited:
Interesting.. I have a 2700x that was boosting to 4.35 but now only hits 4.25 after bios updates. I noticed in bios CPU sits at 1.4 now when it was usually 1.03 or so. Seems like my Voltage isn't throttling down like it used to.
 
In the future, avoid Strix motherboards. They make them as cheap as they can get away with, but charge you for the logo nonetheless. This isn't 1998, being a victim of hype is the consumer's fault and the consumer's fault alone.

The Strix AM4 boards are top-tier VRM across the entire line. I have no idea what you're on about besides trying to be one of the cool kids bagging on Asus. Besides, point me to another AM4 ITX board with two m.2 slots, I'll wait.

Interesting.. I have a 2700x that was boosting to 4.35 but now only hits 4.25 after bios updates. I noticed in bios CPU sits at 1.4 now when it was usually 1.03 or so. Seems like my Voltage isn't throttling down like it used to.

I've seen this mentioned a few times on other forums while looking at this issue. I have never seen mine boost to 4.35 though.
 
The Strix AM4 boards are top-tier VRM across the entire line.

You know less than you think, far less judging by this alone. But since i'm a "cool kid" here to ruin your day, by all means keep buying them then.

Now to something even a "cool kid" like me might help you with, your never reaching 4.35GHz doesn't mean much. There's no magic involved in Precision Boost, you're still limited by arc and lottery just like with any other chip. So if your chip can't take it, it just won't. The mobo however will keep trying, that's why we have presets, so that it adhers to them (my mentioning PBO above); hence my saying check your settings. Your chip may be unable to reach that frequency, but the mobo provides regardless.

A bit of further and final advice and don't get offended again. This too shows you know less than you think you do (it's like saying i keep upping the voltage but nothing changes.. same exact thing, only you now noticed it in the reverse).
Did you see me going on the offensive? Did you perhaps see me reaching to conclusions regarding your character/mentality? Them being totally irrelevant anyway?
Good luck and hope you get this sorted.
 
The Strix AM4 boards are top-tier VRM across the entire line. I have no idea what you're on about besides trying to be one of the cool kids bagging on Asus. Besides, point me to another AM4 ITX board with two m.2 slots, I'll wait.



I've seen this mentioned a few times on other forums while looking at this issue. I have never seen mine boost to 4.35 though.

I have mentioned this on this forum and Overclockers... did you see it elsewhere also? The F2 bios with my Gigabyte Gaming 7 X470 seemed to allow it ot boost to 4.35.. newer AGESA were released and I don't know if that had anything to do with the peak boost but I never see even the claimed 4.3 anymore.
 
You know less than you think, far less judging by this alone. But since i'm a "cool kid" here to ruin your day, by all means keep buying them then.

Now to something even a "cool kid" like me might help you with, your never reaching 4.35GHz doesn't mean much. If your chip can't take it, it just won't. The mobo however will keep trying, that's why we have presets, so that it adhers to them (my mentioning PBO above); hence my saying check your settings. Your chip may be unable to reach that frequency, but the mobo provides regardless.

A bit of further and final advice and don't get offended again. This too shows you know less than you think you do (it's like saying i keep upping the voltage but nothing changes.. same exact thing, only you now noticed it in the reverse).
Did you see me going on the offensive? Did you perhaps see me reaching to conclusions regarding your character/mentality? Them being totally irrelevant anyway?
Good luck and hope you get this sorted.


Mine was hitting 4.35 fine before a bios update. This is not an indicator of the chip just won't do it.. mine did and stopped after the update. Had my board showed up with a newer bios I may have never seen the 4.35. I am on liquid cooling and my temps never go above 60C full load after hours.

Something I should add.. my original F2 bios did not have a PBO option.. this was added into newer BIOS revisions.. I think you are on to something with PBO. Mine is set to auto. I may mess with those settings. This could also be what is holding my voltages higher than normal.. even in bios.
 
You know less than you think, far less judging by this alone. But since i'm a "cool kid" here to ruin your day, by all means keep buying them then.

Now to something even a "cool kid" like me might help you with, your never reaching 4.35GHz doesn't mean much. There's no magic involved in Precision Boost, you're still limited by arc and lottery just like with any other chip. So if your chip can't take it, it just won't. The mobo however will keep trying, that's why we have presets, so that it adhers to them (my mentioning PBO above); hence my saying check your settings. Your chip may be unable to reach that frequency, but the mobo provides regardless.

A bit of further and final advice and don't get offended again. This too shows you know less than you think you do (it's like saying i keep upping the voltage but nothing changes.. same exact thing, only you now noticed it in the reverse).
Did you see me going on the offensive? Did you perhaps see me reaching to conclusions regarding your character/mentality? Them being totally irrelevant anyway?
Good luck and hope you get this sorted.

You can fuck right off with your patronizing shit.
 
Yeah, his reply was a bit much... 165 posts being an ass to a long standing member isn't a great start lol.

Well I solved the mystery. The problem wasn't the board, chip, or BIOS at all. It was NZXT CAM software. I have no fans connected to the NZXT hardware, its strictly for controlling a couple of light strips. I was trying to adjust the brightness and it crashed, lo and behold as soon as it stopped running the fans went static. I rebooted to fans surging, shut the program off and it went back to basically fixed speed again. Control panel wasn't showing any CPU usage for it so I can only speculate what it was doing but I guess I'll have to wait for a non-beta release of the program that supports Zen+ (regular release reads CPU at 95C all the time so they put out a beta).
 
Also something relatively useful I found out in my comparisons, the 1700 ran about -10C across the board from the 1800X in the same system. So instead of idling at 40C (wouldn't be uncommon), idling at 50C seems about to be expected. You have to hit 60C for AMD's throttling algorithm to kick in a way that affects your XFR2 curve with temps, but just use it.

In sum yes the 65W TDP chips are almost always the best choice. I just can't resist trying to make the dumb choices like the 8700K and 2700X work though.

You had a UEFI bug that caused temperatures to be reading 10C off, and the temperature sensors in the 1700 and 1800x report different values. You'd need to tweak SenseMI settings in order to resolve this.
 
You had a UEFI bug that caused temperatures to be reading 10C off, and the temperature sensors in the 1700 and 1800x report different values. You'd need to tweak SenseMI settings in order to resolve this.
I did? I don't think I agree with this. I tried every single UEFI release, including the betas for my board. It seemed to me to be caused by the higher base clocks. I was recording tdie to tdie.
Either way I don't know, if true, by jove AMD, let's get this temp situation straightened out so things are guaranteed to be apples to apples.

You can fuck right off with your patronizing shit.
I'm with LigTasm, but I can be patronizing too, or at least tell it like it really is. Poverty gamers can't afford Asus and don't see the value. I have a ton of parts in my system that would make your average high school kid say "WHY WOULD YOU BUY THAT", like my 960 Pro 1TB (which I regret not getting the 960 Pro 2TB for $1200), because I don't buy bang for buck. I buy exactly what I want (MLC configured memory, which I've used since the advent of SSDs, it's just better than what you have sorry), if it's not perfectly in the middle of the price/performance curve the kids are shatting themselves. I can buy anything on the market, easily. I'm not tweaking out my system I'll use for years, based on pinching pennies. If you want front mounted M.2 on mITX as I do for better temps & cooling, it's Asus or nothing. When I'm tapping my multi-six figure bank accounts that I've built up over the years through my own labor, my wife gets dinners at Fogo de Chao with me whenever she wants at $150 a pop. Stop assuming a mere $100 upcharge makes everyone hesitate for "bad value", study harder, get a job. This mentality in the PC building community, obsessive hand wringing over $15 is sickening, just a bunch of children or adults who don't want to work hard. Yes, some people overpaid for a Titan card. But they got to enjoy a Titan, you didn't, and that's priceless. They'll probably have the next one too, you won't. Keep hating.
Poverty gamers are a blight. Go mow some more lawns.
 
Well I solved the mystery. The problem wasn't the board, chip, or BIOS at all. It was NZXT CAM software. I have no fans connected to the NZXT hardware, its strictly for controlling a couple of light strips. I was trying to adjust the brightness and it crashed, lo and behold as soon as it stopped running the fans went static. I rebooted to fans surging, shut the program off and it went back to basically fixed speed again. Control panel wasn't showing any CPU usage for it so I can only speculate what it was doing but I guess I'll have to wait for a non-beta release of the program that supports Zen+ (regular release reads CPU at 95C all the time so they put out a beta).

Nice Catch! I haven't used the beta release yet.. I also use CAM but I don't use it to control my fans. Wish I could figure out this 4.35 boost issue. I will play wiht it some when I have time. Seems like my voltages never drop below 1.31V when in windows. This may be normal but in bios with first Bios I was seeing 1.03V and now I see 1.4V
 
Nice Catch! I haven't used the beta release yet.. I also use CAM but I don't use it to control my fans. Wish I could figure out this 4.35 boost issue. I will play wiht it some when I have time. Seems like my voltages never drop below 1.31V when in windows. This may be normal but in bios with first Bios I was seeing 1.03V and now I see 1.4V

The ryzen power plan appears bugged to me with the new CPU. With it on, the cores never drop below full speed (4000mhz), with the regular high performance plan it immediately drops to 2200mhz at idle and the voltage drops off as well.
 
The ryzen power plan appears bugged to me with the new CPU. With it on, the cores never drop below full speed (4000mhz), with the regular high performance plan it immediately drops to 2200mhz at idle and the voltage drops off as well.

Ahh, ok... I will give that a look. I would expect it to drop at idle and it just hovers around 4ghz like yours. Thanks for this!
 
The ryzen power plan appears bugged to me with the new CPU. With it on, the cores never drop below full speed (4000mhz), with the regular high performance plan it immediately drops to 2200mhz at idle and the voltage drops off as well.
I thought I read somewhere that you don't want to install or use the Ryzen power plan with Ryzen2. That's a Ryzen1 thing only. I'm not reformatting, but I am going to uninstall all AMD software before swapping stuff out, and running DDU for all things AMD for good measure.

I'm building my 2700X / Asus X470 / C7 Cu + Noctua fan rig rebuild today, so I'll be able to add another datapoint here soon. Just waiting on this to arrive.

upload_2018-5-16_13-51-28.png
 
Last edited:
I thought I read somewhere that you don't want to install or use the Ryzen power plan with Ryzen2. That's a Ryzen1 thing only. I'm not reformatting, but I am going to uninstall all AMD software before swapping stuff out, and running DDU for all things AMD for good measure.

I'm building my 2700X / Asus X470 / C7 Cu + Noctua fan rig rebuild today, so I'll be able to add another datapoint here soon. Just waiting on this to arrive.

View attachment 74364

What the hell is that? Geeze.. I take some time off and so much has changed lol.

I installed the Ryzen chipset drivers off the Ryzen website per info I read. The power plan is easy to change if that is the problem.
 
What the hell is that? Geeze.. I take some time off and so much has changed lol.

I installed the Ryzen chipset drivers off the Ryzen website per info I read. The power plan is easy to change if that is the problem.

I'm going to reinstall the drivers too, but probably use the non-AMD plan. I wish I could remember where I read that. And of course I could be wrong. I always am if you ask my wife, mother, or anyone on the internet.
For the IC graphite pad, it's pretty cool. I thought I'd give it a try. I liked the idea of it because it's a nice perfect IHS coverage everytime. It'll encourage me to do more AM4 upgrades, almost certainly dropping in 7nm Ryzen3 / Zen2 next year because of this thing. I get tired of cleaning/reapplying paste (I'm meticulous).
 
I'm going to reinstall the drivers too, but probably use the non-AMD plan. I wish I could remember where I read that. And of course I could be wrong. I always am if you ask my wife, mother, or anyone on the internet.
For the IC graphite pad, it's pretty cool. I thought I'd give it a try. I liked the idea of it because it's a nice perfect IHS coverage everytime. It'll encourage me to do more AM4 upgrades, almost certainly dropping in 7nm Ryzen3 / Zen2 next year because of this thing. I get tired of cleaning/reapplying paste (I'm meticulous).


Just watched the video.. I use Thermal Grizzley kryonaut. Not sure how it compares. I love how easy the pad would be to apply and re-use. What is the non AMD plan? Just let windows install the chipset drivers? I am surprised that would be better.. not usually the case.
 
It depends how picky a person is going to be. I think you'll be alright but I'm not a huge fan of Kryonaut on ambient temp cooling for multiple reasons, most of the "best pastes" are actually formulated for sub ambient cooling.
I'm a MX4 fan, but a reusable pad within 1C of IC Diamond is overall pretty tough to beat. Obviously people have been squabbling over 1 to 5C temp differences for years in the PC builder space, and will continue to.
I'll be using this 100% IHS coverage graphite pad. :)
 
Well my problem with the CPU not throttling down was the Power plan. I had to move it to 5% and it throttle down now.. I noticed the voltage throttles up higher now when it moves and I am again boosting to 4.35ghz!
 
I've read that you want to use the windows balanced power plan with the 2000 series Ryzen and that you shouldn't use the Ryzen balanced plan but I don't know how the performance plan figures in. With windows balanced I idle in the low thirties at under one volt and all my cores boost to 4.35ghz though obviously not all at the same time.
 
Well my problem with the CPU not throttling down was the Power plan. I had to move it to 5% and it throttle down now.. I noticed the voltage throttles up higher now when it moves and I am again boosting to 4.35ghz!

Sorry for asking but where is that power plan setting ? (I should google it but that is prone to me making error)
 
Sorry for asking but where is that power plan setting ? (I should google it but that is prone to me making error)
I think he's talking about inside the plan, the CPU Usage "Min" and "Max", where the Min for the AMD plan was set to 100% (whether by default or not, I dunno), and so he adjusted it to 5%.
This was actually the solution I was about to post to him, but he had figured it out so (y)

While I know this thread is basically moot, at least for the OP, something else to consider is that adjusting the CPU's Multi may also disable CnQ (Cool 'n' Quiet) on your board. I know it does now on my Titanium as of the more recent BIOS updates. Which, I believe, it can't be re-enabled either so long as you have a custom Multiplier set.
 
I think he's talking about inside the plan, the CPU Usage "Min" and "Max", where the Min for the AMD plan was set to 100% (whether by default or not, I dunno), and so he adjusted it to 5%.
This was actually the solution I was about to post to him, but he had figured it out so (y)

While I know this thread is basically moot, at least for the OP, something else to consider is that adjusting the CPU's Multi may also disable CnQ (Cool 'n' Quiet) on your board. I know it does now on my Titanium as of the more recent BIOS updates. Which, I believe, it can't be re-enabled either so long as you have a custom Multiplier set.

You are correct Sir.. setting to Performance alone did not fix the issue. Setting Balanced may have fixed it but I did not try. I chose performance and then set CPU to minimum %5. All seems better but voltage doesn't seem to ever drop below 1.31V and spikes to almost 1.5V.. acting very similar to overdrive which is set to Auto.. I may disable that in bios and see if voltages drop. I want nice and cool idle temps and low idle voltage.
 
All seems better but voltage doesn't seem to ever drop below 1.31V and spikes to almost 1.5V.. acting very similar to overdrive which is set to Auto.. I may disable that in bios and see if voltages drop. I want nice and cool idle temps and low idle voltage.
I can help but think this is AGESA PinnaclePI 1.0.0.2a related. When I updated my Titaium's BIOS to the latest, hooo boy, the default voltage sure did get crazy. Went from a SummitPI 1.0.0.4a (very old!) BIOS, which while in the BIOS would idle at like high 1.1V to low 1.2V (I think it was like 1.21V), but on this new BIOS is 1.46-1.48V! I initially attributed it to having flashed a modded BIOS, but then I flashed back to 1.74, then updated again but to the completely untouched latest, and it was still applying excessive levels. I was honestly so put off by that, that I manually set it to 1.35V just for testing (which is excessive for my chip by itself), as I wasn't willing to attempt booting into Windows with that :\ I mean I have sufficient enough cooling, but I'd really rather not risk it.

Now, Zen+ I feel a little more confident about, personally. It seems way more intelligent with it's SenseMI, though there may not be any actual (or meaningful) change between Zen and Zen+... but doesn't change the fact that I'd feel more at ease seeing 1.5V from a 2700X than my 1700X, as I know the PB2 will do that.

Though, if the CPU speed is dropping down, my past experience with AMD chips, even my Laptop with Carrizo, when it drops clocks it drops voltage. So if your CPU speed is dropping below its base clock, I can't see a reason for it needing to maintain 1.31V. My laptop's base clock is 2700MHz with a boost of 3800MHz. At full base clock it only requests like 1.2V (1.4V full boost), but like currently it's bouncing between 1.6GHz and 2.4GHz with it's CoreVID at 0.825V and 0.935V, Core Voltage reading basically a static 0.887V.

That being said, I don't know if this works for Zen+, if there was anything major that changed that could break it from supporting it over Zen, but you could always give K17TK a try. I had posted about it long ago, but it's a simple program to change the Boost PStates from within Windows without a restart. I don't know how well Ryzen Master works these days, but in the early days it was pretty worthless as everything required a reboot to apply. Anyways, while the program is in English, the author is Japanese and Google Translate only can do so much lol
THIS is the most recent version, which just fixes something for giving Windows permission to "restart" the program (by the sounds of it... heh).
And THIS is the previous version to that, and I'd say if it works fine for you then use it instead.
Just in case WinRing is still required (since it doesn't come with it), then here's a link to the releases for it (officially it's been long abandoned AFAIK) and the most recent one works fine for me.

Anyways, there are a number of command line arguments you can use by making a shortcut and appending them after .exe (it's a GUI program no worries), so if you end up using it some of those may come in handy. Also, worth noting is that when you open it, it'll automatically apply the settings in the config file. If you make changes and exit, those settings are now the new "default" that are applied when you start, so if those end up being unstable and crash you... be sure to modify or delete the config to reset it. You can also save various different ones and specify to load them with a command line arg, if you need or want.
 
housecat remember to update the IC pad thread with some numbers when you get things together. I'm interested to see what you get since my pad is arriving tomorrow.
 
I always set the Power Profile in Windows 10 Pro to "High Performance". It is also recommended by the Ryzen Master application. I know others have stated they use "Balanced" instead of "High Performance".
I always set the minimum processor state to 5% or less (1%) in the High Performance profile. I have it set to 5% now. Not doing this on the High Performance power profile will result in a CPU that never slows down.

My 2700X (rest of rig specs below) idles at around 0.8125V and all cores at 2200 MHz, however quite often I see voltage spikes of 1.51 - 1.55V that last for a second or two in the Ryzen Master application that I'm using to monitor things. Generally this happens when a single core spikes up in usage for a brief period of time, but not always. Sometimes the voltage spikes to over 1.5V but there is no core speed increase on any of the cores, or at least nothing shows up in Ryzen Master.
 
I always set the Power Profile in Windows 10 Pro to "High Performance". It is also recommended by the Ryzen Master application. I know others have stated they use "Balanced" instead of "High Performance".
I always set the minimum processor state to 5% or less (1%) in the High Performance profile. I have it set to 5% now. Not doing this on the High Performance power profile will result in a CPU that never slows down.

My 2700X (rest of rig specs below) idles at around 0.8125V and all cores at 2200 MHz, however quite often I see voltage spikes of 1.51 - 1.55V that last for a second or two in the Ryzen Master application that I'm using to monitor things. Generally this happens when a single core spikes up in usage for a brief period of time, but not always. Sometimes the voltage spikes to over 1.5V but there is no core speed increase on any of the cores, or at least nothing shows up in Ryzen Master.

the stepping happens faster than the refresh rate of the application, so you won't always see it even though the voltage spikes. it's similar to vega where it's almost impossible to get the data set interval rate low enough on recording apps to pick up the gpu clock changes. there was a guy that did a few recordings of his 2400g playing a bunch of games here on the forum that shows the cpu clocks during the game play and the rate at which it changes is insane.
 
Well I put the loop on since I bought it already, seems to be working good. CPU and GPU both top out around 60C which is fine considering 360mm of rad space in an ITX case with a nearly 300W GPU.

Interestingly, if I turn off the CPB and manually set the CPU at 4.0 ghz, the temps almost never budge off of 40-42C regardless of load. I don't think the X370 BIOS update for the Zen+ is working quite right.

IMG_20180517_223743.jpg
 
Last edited:
Well I put the loop on since I bought it already, seems to be working good. CPU and GPU both top out around 60C which is fine considering 360mm of rad space in an ITX case with a nearly 300W GPU.

Interestingly, if I turn off the CPB and manually set the CPU at 4.0 ghz, the temps almost never budge off of 40-42C regardless of load. I don't think the X370 BIOS update for the Zen+ is working quite right.

View attachment 74746

X470 boards do the same thing.. at least mine does.
 
Interestingly, if I turn off the CPB and manually set the CPU at 4.0 ghz, the temps almost never budge off of 40-42C regardless of load. I don't think the X370 BIOS update for the Zen+ is working quite right.]

I'm slightly of the same opinion on the AGESA on X370 boards being a bit sketchy, even for Classic Ryzen. I'd have to try the various BIOS version in between the v1.74 and the current v1.F0 I'm running now to know if it's a recent thing, but the fact that this newest one by default applies 1.47V to my CPU is a bit weird. It would definitely result in manual overclock temps being much lower.

On the flip side, I may be taking that value with too much initial impression bias, and that realistically the CPU isn't operating at that, making it simply a voltage level that it now can operate at if the XFR/PB determine it needs that much. I honestly can't say because I was too reluctant to let it boot to Windows on Auto Voltage heh
 
Well to add insult to injury the CMOS battery on the board died. Never had one go bad in 20 years of building PC's but I think this one did, and fortunately for me its not just a battery its a stupid contraption wrapped in heat shrink with wires that attach to the board.
 
I've had them go dead, but generally only when the board hasn't been plugged into a powered PSU for a long long long time. So yea that is rather surprising to hear.

The usage of "laptop" battery configuration (wired battery) isn't too surprising in an ITX setup though. They DO make vertical CMOS sockets, so not sure why that wasn't utilized instead heh Oh well :\

Just get a new watch battery for it and solder those wires back on. I would suggest heatshrinking it after, but even taping it up would suffice. While electrical tape is best in theory, in practice it usually doesn't jive well with heating and cooling cycles, which end up making it slide off and leave a lovely gooey film behind. As such, I'd just use the blue painter's tape (scotch tape). It's thick, and pliable, and doesn't move after thermal cycles.
 
Back
Top