• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

RX 9000 series GPU Owners Club

Good for You, I've bought New psu cause of those spikes that even in the reviews where :)
In some games it will also spike for very short time but for some reason not in 3DMark.
 
Thanks for all the help everyone, I reinstalled drivers, reset system bios and reset amd software. I think issue is fixed now, but I don't know which item was causing issue.

The new scores after resetting everything, full stock are now much better,.. I was getting worried thinking my new card was borked,.. :toast:

Stock default settings results.

Time Spy http://www.3dmark.com/spy/55881785

View attachment 399801

Steel Nomad http://www.3dmark.com/sn/6090400

View attachment 399805
Your average temp is pretty high when I compare my run of steel nomad of yesterday with your result.

My run : https://www.3dmark.com/sn/6080541

PL: 100% (340W), UV: -40mV, VRAM @2784 standard timing.

You had 46°C, I had 37°C, it's a very huge difference for a short lived benchmark. Especially since the hotspot temp can sometime be 45°C over edge temp.
 
Your average temp is pretty high when I compare my run of steel nomad of yesterday with your result.

My run : https://www.3dmark.com/sn/6080541

PL: 100% (340W), UV: -40mV, VRAM @2784 standard timing.

You had 46°C, I had 37°C, it's a very huge difference for a short lived benchmark. Especially since the hotspot temp can sometime be 45°C over edge temp.

If I remember correctly the ambient air temp was around 24c, was mid day and I ran several 3DMark bench's in a row while testing.

Perhaps if the ambient air temp was cooler and me only running one short lived benchmark,.. my average temp results would have been lower. :)
 
Your average temp is pretty high when I compare my run of steel nomad of yesterday with your result.

My run : https://www.3dmark.com/sn/6080541

PL: 100% (340W), UV: -40mV, VRAM @2784 standard timing.

You had 46°C, I had 37°C, it's a very huge difference for a short lived benchmark. Especially since the hotspot temp can sometime be 45°C over edge temp.

Here you go,.. adjusted fan curve, cooler ambient air :)




Screenshot 2025-05-18 091216.png
 
Ok, I hate You guys (I'm mad at myself that I did not bought the xfx oc with magnetic fans)
:)
My gpu is showing 42C on steel nomad, and with fans at 100% from case, and 40% on card, and it even lift the calendar that is over 1m from the pc xD
1747575498427.png
 
Ok, I hate You guys (I'm mad at myself that I did not bought the xfx oc with magnetic fans)
:)
My gpu is showing 42C on steel nomad, and with fans at 100% from case, and 40% on card, and it even lift the calendar that is over 1m from the pc xD
View attachment 400211

The Mercury OC has a vapor chamber cooler with 8-pipes,.. works very well with a low RPM fan curve,.. I'm still testing.
 
Last edited:
Gigabyte has vapor chamber too, can You add the gpuz stats from steel nomad?
 
Last edited:
Could You add max temps or set in the gpuz or set in the settings sensor display mode highest, and reset before next test, and than add with highest temps on card ? :)

1747584210618.png
For example steel nomad 90% (300W) power limit, -80mV, custom fan curve.
Wynik.JPG
 
Last edited:
Sorry,.. still learning this GPU TWEAK III interface,.. this is fans @ 50%


Screenshot 2025-05-18 120406.png


Screenshot 2025-05-18 121231.png
 
Last edited:
Not a problem, everyone is learning :). and we are here to get a lot of info :D
So XFX is really nice card, and gigabyte is not a shabby one too :D
 
Sorry,.. still learning this GPU TWEAK III interface,.. this is fans @ 50%
Adrenaline offers far better GPU tuning settings than this GPU TWEAK III, what's the point of using it?
 
Adrenaline offers far better GPU tuning settings than this GPU TWEAK III, what's the point of using it?
Adrenaline is bloated with stuff that is not needed, and for example in My pc, adrenaline was crashing out even on stock settings.
With moreclocktool I've got -80mV stable in path of exile 2.
So Maybe GPU TWEAK III could yield better results, Cause I see there -102mV and for me -80 is stable, -90 is problematic, and -100 is shutting down pc (-90 and -100 was tested under linux, -80 was tested on linux and windows 10 ltsc)
 
Adrenaline is bloated with stuff that is not needed, and for example in My pc, adrenaline was crashing out even on stock settings.
With moreclocktool I've got -80mV stable in path of exile 2.
So Maybe GPU TWEAK III could yield better results, Cause I see there -102mV and for me -80 is stable, -90 is problematic, and -100 is shutting down pc (-90 and -100 was tested under linux, -80 was tested on linux and windows 10 ltsc)

That's interesting but strange. Did the same voltage offset give the same boost clock speed for Adrenaline vs MoreClockTool. I thought the same clock and same voltage should be stable no matter which application set it.
 
Adrenaline is bloated with stuff that is not needed, and for example in My pc, adrenaline was crashing out even on stock settings.
With moreclocktool I've got -80mV stable in path of exile 2.
So Maybe GPU TWEAK III could yield better results, Cause I see there -102mV and for me -80 is stable, -90 is problematic, and -100 is shutting down pc (-90 and -100 was tested under linux, -80 was tested on linux and windows 10 ltsc)
It's very strange to crash the stock settings, there's probably another problem with the windows.

With third party tools you only get a few points more in some benchmarks, nothing more, and lose access to picture quality settings and detailed GPU overclock settings.
The only thing I do is stop Adrenaline CPU control via Microsoft Autoruns, where Ryzen Master should be disabled because it broke my CPU overclock or I need every time to choose only GPU control when save/load settings.
 
had a undervolt oc on my 7900 xt for a long time now, just using drivers to oc, no issues here. something else is going on its 9000 series only having issues i dunno
 
It's very strange to crash the stock settings, there's probably another problem with the windows.
It was Windows 11 ltsc 24h2 or something like that and 25.3.1 adrenaline. And on windows it was unstable as f...., but on linux card was stable as f... :D
Moreclocktool has some settings from the adrenaline in it.
 
Did anyone notice lower benchmark scores with latest AMD driver 25.5.1? I'm getting roughly 150-200 lower scores with same settings.
Before, I was able to score 7200+ score with out of the box settings. With the new driver, I get 7060-7080 using same settings.
 
Did anyone notice lower benchmark scores with latest AMD driver 25.5.1? I'm getting roughly 150-200 lower scores with same settings.
Before, I was able to score 7200+ score with out of the box settings. With the new driver, I get 7060-7080 using same settings.
I can test later today.
 
Did anyone notice lower benchmark scores with latest AMD driver 25.5.1? I'm getting roughly 150-200 lower scores with same settings.
Before, I was able to score 7200+ score with out of the box settings. With the new driver, I get 7060-7080 using same settings.
Which RX 9070 XT are you using ?

In steel nomad i'm getting the same score but in unigine superposition 25.5.1 is slightly faster than old drivers.

25.5.1 performs better overall but also undervolting potential is slightly dropped because of that. They are squeezing more juice out of newer drivers.
 
Last edited:
Adrenaline offers far better GPU tuning settings than this GPU TWEAK III, what's the point of using it?

I was having issues overclocking card, crashes, drivers not loading, failed 3DMark result's,... I have been using Adrenaline since day one.

I switched to GPU Tweak III and all the issues went away, I have not changed anything else.

GPU Tweak III offers far better tuning experience on my system.

Everyone's hardware is different,.. if your OC is unstable you need to keep testing, making changes,.. find what works best for you and your hardware.
 
Last edited:
Which RX 9070 XT are you using ?

In steel nomad i'm getting the same score but in unigine superposition 25.5.1 is slightly faster than old drivers.

25.5.1 performs better overall but also undervolting potential is slightly dropped because of that. They are squeezing more juice out of newer drivers.

Assus 9070 XT TUF
 
arma.gif
I like this card, I was playing arma 3, smooth with no problems (of course i don't count the arma bugs :P ) and temps are great :D
But this 3411 mhz on core is strange :D
 
Did anyone notice lower benchmark scores with latest AMD driver 25.5.1? I'm getting roughly 150-200 lower scores with same settings.
Before, I was able to score 7200+ score with out of the box settings. With the new driver, I get 7060-7080 using same settings.

25.5.1, no difference here, just the same result as before.

 
Back
Top