[SOLVED] Voltage "issue" on ASUS TUF Gaming OC 6900 xt

[David0ne86]

Hi guys
I'm writing here because I have a minor issue with my newly acquired asus tuf 6900xt.
Basically no matter what voltage setting i give the card either with the adrenalin software or Asus proprietary software (gpu tweak 2) it keeps bouncing up and down instead of staying at a fixed value as I've set (i would like to have 1100mV instead of the ridicolous at stock 1175).
I even made a short clip to show you guys what I'm talking about. If you watch the video you'll notice the slider being at 1100mV but the reading keeps jumping up and down (and this is not an adrenalin bug, because even gpu z shows such back and forth jumps).
I tried using either adrenalin software only while gpu tweak 2 being uninstalled, and having gpu tweak 2 only hile adreaniln was not installed (i installed the drivers via device manager).

Any of you have any idea what the issue could be? I come from a 5700xt and had none of this issue (altho that card gave me way more problems especially at launch lol) and ofc before installing the 6900xt i used ddu via safe boot to properly clean the drivers (i even checked the little "perform factory reset" box when installing the new amd drivers.
Here's the link to the clip

View: https://youtu.be/_nffIsCAmqg


Hope someone can give me a possible reason on why the card is doing that. It's not really a major issue per se, but since i paid good money for it i wouldn't want to compromise and since the card CAN BE undervolted i would like to do so to further shave some degrees off my card.

 

[Ralston18]

Update your post to include full system hardware specs and OS information.

Include PSU: make, model, wattage, age, condition.

Instead of DDU, go to the manufacturer's website to find and manually download the drivers.

Reinstall and reconfigure.

 

[David0ne86]

Hey, first and foremost, thank you for replying. Secondly maybe i expressed myself wrong but i didn't use ddu to INSTALL my driver (i didn't even know that was possible) i simply used it to uninstall previous drivers since i come from a 5700xt. So i used DDU in safe mode to remove the old drivers (after i ofc uninstalled the radeon suite via the control panel in windows). I also disabled windows to automatically check for drivers to update, so that's off the equation too. Drivers were downloaded by amd official website.
Besides that, you're right, i should've been more thorough with my pc specs... so here we go:

My OS is
Edition Windows 10 Home
Version 20H2
Installed on ‎1/‎16/‎2021
OS build 19042.867
Experience Windows Feature Experience Pack 120.2212.551.0 (i pretty much updated it on tuesday this week, so no other updates are out yet)

CPU it's a ryzen 7 5800x stock, motherboard is a msi mag tomahawk x570 bios 7C84v15, ram are 32gb gskills trident z neo 3600 mhz cl16 ddr4 (double kit of 2*8 gb, making sure they're IDENTICAL, i used cpuz to make sure) and my psu is a corsair hxi 1000w, and it's practically brand new, i bought it for this build 3 months ago. As drives i use 2 sabrent rocket nvme and one ssd sata (crucial mx1000). Also in my original post i forgot to mention that i tried out two versions of the amd drivers, the 21.2.3 (the one i show in the youtube video i posted and the most current one) and the previous 21.2.2. Still getting the same jumps in voltages. Please feel free to let me know if i need to be more specific on anything else

 

[helper800]

What exactly do you mean when you say "(i installed the drivers via device manager)."? Because this could be a huge issue. Something to consider is that setting the voltage lower may not limit it, but be a baseline for boosting. Can you try setting it lower than 1100mv to see if the card actually reduces voltage at all?

 

[David0ne86]

Hey, thank you for taking your time

By installing via device manager i meant to say that i installed the amd drivers using device manager as shown here

View: https://www.youtube.com/watch?v=IhBLqTVxcBM&


It's a pretty common thing to do with amd cards if you want to avoid the proprietary software as it's the only way, unlike nvidia that lets you choose if you want to install just the display drivers or the whole suit in its installer. Nothing wrong can come from that. Besides, i did try both versions, first with the normal installer (henche why you can see in the youtube video of my original post im using amd proprietary software) and after seeing no matter the settings kept jumping up and down, i did an installation via device manager to then try use the asus proprietary software for the card (gpu tweak 2) to further minimize the chance both softwares would override eachothers.

And trust me, i set the voltages even to the minimum possible (which is 775mV) but no matter what, the jumps up (and well at that point not down since it's impossible for the card to go below that voltage) keep happening. You have to see that slider as an hard capper. Pretty much it's like saying to the card via the amd software to never go beyond whatever setting you put on the voltage slider (in my case 1100mV). It always worked that way even when i owned a 5700xt and i was pretty much forced to undervolt it otherwise that card would go up to 110 degrees on the hotspot since it had insane stock voltages (1200mV).

Im starting to believe it's just a driver issue being still immature due to the fact the card is so new (it launched not even 2 months ago) so it'd be understandable if the drivers would still a bit iffy, especially amd ones which have been known to always be on the unstable side (had amd cards pretty much all my life except from 2014 to 2018 as i owned a gtx 1070). So im pretty much used to these hiccups, especially at launch. I was hoping someone here would own the same card as mine and could share their experience, but apparently not to many 6900xt tuf owners.

 

[helper800]

So one thing to note is that the model of card you have (be it Asus, Sapphire, Asrock, MSI, et cetera) shouldn't have any impact on the software's ability to change settings for the cards. As long as it is a 6900XT it should be able to be manipulated by any software that has access. Have you tried parsing through any on gamer's nexus youtube channels on overclocking the 6900xt and 6800XT cards? I believe the one they did with the 6800XT they ended up undervolting the card to 1100mv for the best performance. That video might shed some light on the way they were able to limit the cards voltage.

View: https://www.youtube.com/watch?v=41CiQDXAh0U


Notice I was saying "should" a lot. That is no mistake. I am referring to typical behavior and not what I believe you are experiencing which seems to be atypical.

 

[David0ne86]

Yup. Tried adrenalin (amd proprietary software), gpu tweak 2 (asus proprietary) and msi afterburner (3rd party). Mind you none of them installed together so they could not interfere. And yet, here i am, still seeing my voltages bounce around lol.

I ofc contacted asus already and they gave me the usual pr <Mod Edit> statement "those settings are within normal range blabla" yeah thanks i know they are. But i want to get the best performance out of my card. And having an insane 1.75 v on it doesn't help. <Mod Edit>, that's almost cpu voltage we are talking about. There's no reason for a gpu to have such insane voltages, especially amd card when it's renown that their architecture works better as temperature decrease, and that doesn't help my temp decrease. Are they within normal range? Sure 70 on the core and 90 degrees on the junction are well within spec but they are definitely not optimal.
That said i watched plenty OCing vid and everyone seems just fine by moving the slider down like i did. Tho noone was using a tuf gaming. They were either reference or xfx cards.

I'm between a rock and a hard place, because i could use amazon's send back policy, i have 30 days. But it'd be sad to see they fix this issue in the next or couple next drivers releases and i gave away a great card (besides, ultimately the card still performs great, i just wish I could squeeze every bit of optimization possible).. So I'm really thorn on what to do.

 

[teoo]

can anybody please tell me the idle temperature of 6900 xt, especially in tuf 6900 xt, in a full tower case with only cpu fan working? i dont use any airducts inside the tower. if you undervolt this card or limit max frequency or limit fps (will this work to decrease temperature?), will the idle temperature decrease? if yes by how much? answers would be appriciated, thanks

 

[David0ne86]

can anybody please tell me the idle temperature of 6900 xt, especially in tuf 6900 xt, in a full tower case with only cpu fan working? i dont use any airducts inside the tower. if you undervolt this card or limit max frequency or limit fps (will this work to decrease temperature?), will the idle temperature decrease? if yes by how much? answers would be appriciated, thanks

i can't say for sure since mine is in a very well aired case (corsair 5000d airflow) and temps under stress (especially with last gen games AKA Cyberpunk2077) go up to 70 degrees for the core and 90ish for the junction temp with the fans set up to speed at like 60%. I wouldn't advise to put such powerful card in a non well ventilated case tbh. You can expect atleast +5/6 degrees in my opinion with a non airflow case and no fans. Why would you limit fps if you spent 1500+ euro/dollars on a card. Just buy a less powerful card to begin with. In idle im assuming you can expect normal temperature. It has a zero rpm feature, usually temps are at around 42/43 at desktop, 52/53 doing light work (youtube). So in your case i'd expect idle temps at around 48/50 degrees, light loads 60/63 degrees, gaming 75/78 degrees on the core, add +5/10 degrees on the junction temperature in idle or light work (so around 70/75 degrees) and +15/20 while gaming (so atleast 95/105 degrees).

Can i ask you why you don't have any fans in your pc? That's a bad thing to begin with, you should atleast have 1 infront and one on the rear. ATLEAST 1 on the rear bear minimum to get some hot air out of the case ffs. To me your train of thought really doesn't make any sense.

 

[teoo]

thanks for the reply, i am actually after silence, i love silence, i have hafx case, 230 front, 200 top and side, 140 mm rear fan, which kicks in with ai suite 3 when necesarry. so when i browse internet i dont hear anything from my case, i wanna keep it that way during light work (web browsing). for the fps, normal human eye cannot see any difference above 90 fps , so dont buy in the marketing hype. while gaming if i dont want too much heat, limiting fps should probably decrease my gaming temperatures, i am not sure though, give me your opinions please. where i live 6800 xt and 6900 xt are have very similar prices, so i might as well buy 6900 xt. i intend to buy 3D set in the future to use with my gpu, and when i feel like playing on a 4k 65 inch led tv i could easly remove my undervolt. i also live in a very hot climate. i am looking for ways to keep my card cool when my computer is idle, so i assume undervolting or limiting max frequency should help me , i am not sure which will help the most, and if i ever wanna feel the power of my card i can always put it back to the original settings. i just want my card to give the minimum amount of heat when i browse web, which i do a lot, while running only my cpu fan, i have one fan on my cpu at the moment but i can add one more . thanks

 

[David0ne86]

When the card is in idle it underclocks automatically, it doesn't blasts 2.5 ghz in idle lol. My card goes from 3 to 5 mhz while browsing (like right now writing this, it fluctuates between those values) and goes up between 50 and 500 mhz when i watch a 4k youtube video. And as i said, you have the 0 rpm feature, so the card won't even start its fans when in idle or while watching youtube. And let me tell you, i can clearly see the difference between 90 and 120 fps now that i know what 120fps look like while gaming. Above 120 fps yeah, it's impercettible, but it's not just marketing. I guess you'll see for yourself if and when you'll purchase it. Best of luck.

 

[Karadjgne]

Many fans, especially the better fans from Noctua, Arctic, Noiseblocker etc are effectively dead silent at below 800-900rpm. Running them at @ 500-600rpm will still move air, enough that it can somewhat drastically lower temps inside the case without having that exhaust fan kicking in/out. It'll also move enough air to lower the gpu/cpu need for cooling, which further adds to quietness.

Silent or very close is much preferable as a slight variation to having quiet/WHIRR/quiet/WHIRR... as passive/eco fans kick on/off.

There's multiple voltages when it comes to cpus/gpus. Some of which you cannot control with just 1 setting. VID and vcore is a common pairing, that's the voltage demanded from the VRM's, and the voltage demanded by the cpu. You can set both static and still see a variance, because LLC is a variable voltage added on top of any settings to maintain stability with vdroop. That's going to change with boost levels.

The 6 series gpus is a different architecture to the 5 series gpus, same as Zen3 Ryzens are different to Zen2 Ryzens. So their boost characteristics, voltage use, current use, boosts and changes are different. Also consider the vram settings, nvidia calls it ReBar, which is where the gpu now has full continuous access to all vram instead of portioned by application. That's goin to change memory controller stuff, which can affect boosts and usage, and consequently voltages. 6 series gpus have it native, it's driver added for 5 series.

What it boils down to is you trying to force certain restrictions on the pc according to your wants/desires, and the pc is going to give you the proverbial middle finger and do what it needs to do to function regardless of your wishes or wants. And you cannot compromise with a machine on minimums, cross that line and it just shuts down.

 

[teoo]

i know when the card is idle there shouldnt be any loads on it, i will probably have around 42 celcius on idle, i am just looking for a way to decrease idle temperature by playing with voltage (i hope this does effect the idle temp) and decreasing max frequency (this probably wouldnt effect the idle temp). can people give me their idle temperatures for this card on light load like web browsing prefably while only cpu fan is working in a full tower case? is there anybody who undervolted the card also checked for idle temperature whether it has any effect on it? if 6800 xt has better idle temperature on light load, i could go for it also, if anybody have any ideas, enlighten me please, thanks in advance

 

[Karadjgne]

Idle is a reflection of airflow. You could take 2 identical cases, cards, setups, voltages etc and see 2 very different idle temps if one case has decent airflow and the other does not.

So no, you'll not get a comparison that's justified since you have no airflow and almost any other pc user will, full tower or mITX.

It's impossible to cool an object to below ambient temps by mechanical means. Your cpu/gpu ambient air is the air inside the case, not the air inside the room. So if the air in your case is sitting at 50°C, then the very minimum your cpu/gpu could ever possibly be is 50.1°C. That's why airflow is so important to idle temps, because normally the cooling would be able to get you 30's, but since the cpu/gpu/ram/chipsets etc are all dumping heat into the case, which is not getting shoved out and replaced by room temperature air, you'll not be able to cool the cpu/gpu to below case air temps.

And once you apply any load, you are starting at 50°C and then going up from there, the cpu and gpu using 50°C air or more to try and cool something instead of using outside 22°C air instead.

When you get overheated, what works better, a cool breeze or a hot one? Get rid of the heat in your case and ALL temps drop. If the cpu temp/gpu temp is lower, any fans will not need to be as aggressive, = lower speed fans, = less noise.

Or no noise at all.

 

[teoo]

i use Prolimatech Genesis Processor cooler and Phanteks PH-F140HP_BK2 Processor fan. i have rog maximus x hero mother board and when i select silent profile from fan expert 4, all the case fans on my hafx is stopping, right now my cpu fan is 457 rpm,my cpu temp is 35 C, motherboard 31 C, PCH 44 C, i assume these are not bad for an aircooled 8700k. so the air inside my case cannot be 50 C. i am sure there has to be more people like me, at least i hope so, so if anybody out there could help me, it would be much appriciated

 

[Jacozeelie]

I hate a hot pc and also dont like fan noise. My gtx1650 super oc had a very very loud fan which will run full speed when gaming, i tried many configarations of fans and stuff. Undervolted gpu, caused pc to freeze or crash. Tuned fan curves, cause pc to restart. Eventualy i installed a 120mm aio on the gpu to make it shut up

 

[Karadjgne]

And what are the temps after sitting idle for a half hour, an hour. Heat doesn't build instantly.

With the silent preset, there's also a switch that shuts the fans off below a certain temp, or the fans have yet to be run through the fanXperts optimization, so may not be getting the right voltage to actually run, only getting enough when temps go up and FanXpert's fan curve reaches a higher point.

The Genesis is a funky cooler, it's L shaped design would lead one to believe it's a better design than most, but it's really not. It's very average as far as a heatsink goes in the supposed upper class. But a lot of its performance is dependent on the fans used since normally it doesn't come with fans. If you use the high grade super quiet Prolimatech slim 140's, they run 300-1000 rpm, but don't have much in the way of cfm or sp compared to other 140mm fans, so are seriously quiet, but don't perform above average. Use of some Noctua A14x25 would see far better performance, resulting in lower temps all around, but at the cost of a little more noise at higher rpms.

The Genesis isn't a bad cooler at all, just a bit expensive when compared to other similar weightclass coolers that actually do come with fans.

No, temps as reported aren't bad at all, but consider your cpu fan is already at almost 50% rpm, at cpu idle, so by the time you fully load (or close) that cpu, you will be hitting 100% fan long before reaching full load temps. Lack of airflow through the case is forcing the cpu fan to work harder just to keep that 35°C, which is right where your case temp is at currently. After an hour of sitting and letting case temps acclimate and saturate, your rear case fan will be on just to pull the heat out in a negative pressure system. I'm guessing cpu temps will be closer to 40°C at that point.

Gpu as well, and undervolting isn't going to cure that as volts at idle are higher than volts at load regardless of where you try and set them. At idle, you aren't using power. Power = volts x amps and at high loads its the amperage used that's responsible for all the heat, voltage is the multiplier of the amperage. At idle, you aren't using high current flow, so you get extreme low power use, very little heat, not even enough to overcome case heat outputs. On a test bench, that same gpu at idle would be running a hair over room ambient temps, but with your case sitting at 35° to start with, it's going to be a hair higher than that.

If room ambient is 22°C, cpu/gpu in a case with good airflow will run 8-10°C higher, about 2°C warmer than the case air. In a case without good airflow, that number has nowhere to go but up.

 

[bignastyid]

teoo, please create your own thread, don't hijack other users threads.

 

[teoo]

thanks for your comments karadjgne. As long as i dont play a game my temps stay around the same all day without running a single case fan. my question is completely relevant with undervolting, because as far as i believe that has to do something with idle temperature, thats why i continued from here, so people undervolted can report whether idle temperature goes below or not. thanks

 

[Karadjgne]

Actual temperature or voltage aside, the heat generated by a gpu at idle is really small, a 3080 for instance only uses @ 9w ±. Normally you'd have fans moving air, even a 50cfm pair of 120mm fans running at idle speeds will move @ 10cfm or 1cubic foot of air in @ 6 seconds. Since your average case is somewhere around 1 cubic foot in volume, that's total air replenishment in 6 seconds. 9w from a gpu doesn't stand a chance of seriously affecting anything at that rate.

As far as undervolting goes, if a gpu at idle uses 9w and voltage is 1.5v on the core, that's a total current of 6A. P=VxA. So even if you undervolted like a bandit and dropped voltage to 1.0v, all you'd do is succeed in raising the amperage to 9A.
Because the card minimum ability to function requires 9w. As is, undervolting is generally in the 10mV range, which isn't going to change anything. You'd end up possibly with an unstable card at high loads or loss of boost values since the card now doesn't have sufficient voltage and is maxing out on current, = loss of power.

Trying to close a wound with duct-tape and justifying it with 'well it stopped the bleeding didn't it' isn't going to change the fact you should have gotten stitches. Undetvolting a gpu to lower temps at idle being duct-tape, turning on the fans and moving the air being stitches.

I have 4x Noctua fans, all on radiators. I had to purchase a RGB mouse because the pc even under a decent load is so silent, it's silent and my wife kept turning the pc off, thinking she was turning it on. I had to train her to look at the mouse first, if the light was on, so was the pc. (monitors shut off after 5 mins at idle).