Pretty sure I'm not the only one having this problem.
37W power usage at idle. when using 2 or more monitors.or high refresh rate.
thanks, and pushing us to Nvidia
Same here on 6800. Video playback with 31Watt and 2000mhz when on nvidia there is max 700mhz (400 typicaly)
this is NORMAL behaviour when your Monitors dont use "LCD Standard" or "LCD native"! (will be same for Nvidia cards aswell)
if they use "LCD reduced" or even lower custom settings your VRAM is forced to a higher powerstate!
depening on how your Monitors are set by factory it MAYBE reduces clocks when you lower both screens to 60Hz - as this will have a different powerstate for VRAM.
you can try to "fix" the non-existing issue with OC on your Monitor:
1. use CRU to bring your "Blanking" (and maybe even "Front porch") near or up to "LCD Standard" settings
2. this will raise your Pixelclock and therefore mess with FreeSync; it maybe will be even unsupported by your Panel with resulting in a black screen or artifacts
About "NORMAL behavior" yeah I know But my display compatible with AMD FreeSync Premium. But N cards just work fine.
reduced screen to 60 hz --- works but I guess that's not the solution when you purchase an RX 6000's GPU and FreeSync monitors.
@smilelylifeday
it is simply a standard - if your Monitor doesnt use "LCD standard" or "LCD native" your GPU will need a higher VRAM powerstate to avoid errors
like my RX 580 (and all GPUs since i own my AOC G2460PF) also runs at "3D load" VRAM powerstate all the time (aka 1750MHz or 2000MHz) - because my AOC G2460PF uses lower settings than even "LCD reduced" - this forces my GPU to use higher VRAM powerstate for stability
btw - this also happens with most Nvidia GPUs
you can "fix" this with CRU - but that would OC your Monitor (and i dont recommend that)
ps 2000MHz is still stock - and as there is not much load they wont use much power and wont generate heat...
btw - if you have 2x the same monitor it sometimes even work if both are "LCD reduce" - its some kind of math behind it, but i am to lazy to calculate that
my G2460PF and G2460VQ6 always ramp VRAM clocks to highest powerstate - with every GPU out there - as it HAS to be ;)
Not normal for me. On previous gpu it wasn't normal to ramp up vram to max, it was idling on 400-700 depending on how heavy video I have thrown at it (700 while 2k 60fps quality youtube video) and it wasn't working as bread warmer with 60+ temperature just because it was playing 720p video. So please, don't make bug a feature.
Check my thread with more details Re: RX6800 overheating while watching media on pas... - AMD Community
don't make normal things bug (or at least a insignificant bug as something that waste your life) ... i think explaination as been given, you don't care about what people are telling
Once again, I'm not a newb in those things, I have tested few computers before I was sure something is not right. In my job I have couple computers, with nvidia and amd gpus, rx570, rx580, gtx970, gtx1660, all of them are not overheating by just freaking watching video. Even during using 3d cad programs they are barely using fans to cool down. So once again, amd fanboy, this is a bug.
but you are acting like a newsbee ... did you check your task manager , i think the answer is there generally
did you check the spec of you gpu ? people says it's ok maybe there is a reaon why ... thing can change , not because it was like this before that it's has to be like that now , at least I don't see anything burning , your not speaking about problem with fans or i don't know what other things could indicates something REALLY strange
The current 6000 series users are too rare, leading AMD to think that everything is normal
yeah...guess people bought 6800 series for 1080P 60hz monitor. or mining.
dude - its not a bug...
AND I TOLD YOU WHY ITS NOT A BUG...
BUT if your Graphics Card is at 60°C idle than you have a different problem with it...
1. run DDU in safe-mode to clean old drivers
2. install 20.12.1
try with just 1 screen - check if VRAM clocks down then
0 rpm ... everything is normal ... if it was 80 degrees it would be just weird , but you're speaking about something you seem to don't understand and do not want to understand, you just want to complain or what .. i did tell you to watch your gpu spec , but you didn't
On literally every card I have tested and owned so far (like 10) most of them had zero fan mode and NONE of them was reaching even 50C while watching videos. Please, stop being so hard AMD fanboys and saying "it's normal"
@BrainsCollector
... AS I SAID - ITS NOT A BUG BUT AN INDUSTRY STANDARD that the VRAM clocks to a higher powerstate when using Monitors with "LCD reduced" or lower...
ps my RX 580 also has 0 rpm fans - with my AOC G2460PF (lower than "LCD reduced") and my AOC G2460VQ6 i also have 2000MHz on VRAM constantly - my temps are 40-45°C while normal browsing or watching a video...
IF HIS CARD GETS 60°C OR HIGHER than there is a different problem than just his higher VRAM clocks!
1440P@60 Hz= normal 43 C on Youtube HD' and VRAM CLOCK speed will down like"normal"
and total GPU only cost less then 10W for normal task
GPU will use more the 35W @ idle when you have 1 or more monitors .
not everyone has knowledge of "LCD reduced" (WTF?) should be AMD or Monitors company to fix modify number's
all I feel is FREESYNC? Premium?
I'm using native resolution. Please, don't be hard fanboy. And I'm not using monitor but 4k TV. AND AGAIN PREVIOUS AND ANY OTHER CARD DON'T HAVE SUCH PROBLEMS. It's not hardware related, nor setup or anything. Pure drivers BUG.
While normal web browsing it's reaching 300-1000mhz clock on memory with 50-300mhz on core. I was trying various fixes like adding chrome/edge to power saving list, enabling gpu/pcie power optimization, max power saving priority. Everything that could be worked around.
Here you go mr it's not a bug it's feature. 3 different computers with different monitoring programs, all of them showing minimum memory frequency during normal media browsing
gtx 1660, rx570, rx580, gtx 970
if you think it's a bug you can report it to amd , at the top right of adrenalin, see what they tell you
we are not here to resolve bug , this is no amd customer service forum
you have our view , can't tell you more , even if you insist 100 times more ... if the card doesn't please you , there is no point on insisting like you do, sell it : i will buy it ... lol
Community is about helping not licking each other d* and praising. But seems this one is about that
Why do you even need community forum if it's only about being fanboys ?
if you think it's a bug you can report it to amd , at the top right of adrenalin, see what they tell you
Yes... best solution is just do nothing
agree with you its a bug. only happend when i using RX 6800 XT GPU . everything fine with GTX 1080 RTX 3070
The solution? They just kept talking about CRU, but didn’t say how to do it.
@smilelylifeday
IT IS NOT A BUG! if it works fine with Nvidia GPUs they dont comply to standards!
as i said - it has to do with blanking time and front-porch - it only happens if they use "LCD reduced" or lower settings for the panel.
ps i have the same behaviour with my RX 580, my RX 480, my RX 460, my HD7970 with my AOC G2460PF and AOC G2460VQ6
hmmm
LCD reduced" or lower settings for the panel
So we have to lower the setting if you own a 6800 XT + nice monitor.
come on...And none of one Youtubers talking about this
@smilelylifeday
actually its the other way around xD
"LCD Reduced" is lower settings than "LCD Standard" - and its "LCD Standard" that allows VRAM to clock down
those "nice" monitors are simply going "below spec" in LCD Standards and therefore this error happens
and if the Monitor producer decides to use a Panel in "LCD Reduced" or lower than it will be this for ANY GPU (either AMD or Nvidia or Intel) - so it has nothing to do with AMD.
to be clear: "LCD reduced" is the panel setting. its nothing the GPU does. its caused by the hardware in the panel
Sorry, I really don’t understand. So is this a graphics card problem or a monitor??
@smilelylifeday
its a monitor problem - because the monitor producer (AOC, BenQ, Eizo etc) dont use "LCD native" as standard but "LCD reduced" or lower for their Panel to achieve the high refresh rates and "overdrive" (low response-times)
we could argue that AMD shouldnt be that strict about this standard (as Nvidia often is also not that strict) - but you can have problems when your VRAM does clock down in these cases...
So, it's not monitor manufacturers problem, they made their products long before this gpu was released
@BrainsCollector
WRONG! they dont use the LCD Standard - thats why its not working!
I am also having this issue, but I only have one 75hz monitor and not two. Is it still normal for it to happen with only one monitor, or is there something going on that I should fix? My idle temps used to be around 32 Celsius and now is at 52 Celsius.
my gpu is asus rx 5500 xt 8gb
There is a solution that requires use of an external software for monitor tweaks. Can't remember where I saw it.
But a simpler solution is to set your monitor's refresh rate to 60 Hz if you have a 75 Hz monitor and 120 Hz if you have a 144 Hz monitor. I don't quite know why, but the drivers have a gripe with 75 and 144 mainly.
Also, you need to do this only for desktop. No need to reduce refresh rate in games.
Yes, this can be a way for some people too. But for example with my ViewSonic VP2768 monitor which has a frequency 60 Hz (Quad HD 2560×1440, IPS, 16:9, 14ms, 14bit), lowering the frequency is impossible and therefore I think a better solution is to use an older driver...
2000 vram usage and 37W are not related i think, my vram usage is usually 2000mhz too
nothing alarming about 2000mhz
you can go to nvidia , see if problems are the same or if you find other bugs, i come from nvidia (1080ti) , had lot of little bugs too, nothing magical here (for example my card wouldn't get above some frequencies in games when multiple display etc etc i can do a list of problems i had if you need, just check nvidia forums you see people like you lol)
hey . topics talking about VRAM CLOCK running at max speed all the time . Not the vram usage :)
btw I came from 1080 too
@smilelylifeday...yeah i was talking about vram clock showing 2000mhz usage, sorry for not being explicit enough ...
Enjoy AMD GPU lol. 5700 XT has the same problems. Haven't fixed it for two years. I don't think they have a clue how to approach this, they barely fixed most of black screens[might be also due to wifi interference/emi] so they pray the jury rigged code won't fail when messing with it.
So is there just no fix? Do I just ignore it and let it happen? I’m honestly confused as to why this is even happening in the first place. This problem has been happening for me with every recent driver. My gpu is asus rx 5500 xt.
After installing the older version 20.10.1, idle memory frequencies and temperatures are fine.
There is no fix yet, although they know about this bug, you can find it in release notes / know issues "Radeon performance metrics and logging features may intermittently report extremely high and incorrect memory clock values."