It is probably a very good test setup as it also removes teh Windows driver scenario. I would give it a try. Personally I've never heard of Vetroo supplies.
-
Notice regarding image hosts and attachments. See here.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
The Radeon RX9070XT / RX9070 Owners Thread
- Thread starter CAT-THE-FIFTH
- Start date
Interesting tweaked my 9070XT today when it came and I'm getting crappy GPU timespy scores for some reason barely quicker than my old 6990XT... around 23,300 points, GPU over 3000mhz pretty much all bench and memory set to 2600 and max hotspots 83C... 10% extra power... gonna have to do my home work was expecting way higher than this... Played BF6 thought ULTRA and was lush
I didn't do any diagnostics yet, but my score dropped as well.
From 26676 in February to 23514 now, with same hardware, graphic score went down from 31333 to 27298.
Funny enough I have the same CPU as you.
But I have 64Gb of ram 3600mhz instead of 128Gb 3200mhz and I'm using a X570 mb
My last score: https://www.3dmark.com/spy/66412642
Last edited:
Not that I can see everything running like a dream but what I don't wanna be doing is running blind with an anomaly and throwing away free performance in other games etc and it's tough now, as if I had an exmaplantion on this one, I'm happy but not knowing is doing my head in... is something amiss really.
I might run memory at 3600 in my loser timings then and see if this is a potential stumbling block? Don't see anything else though as a major blocker.
Yeah that's totally understandable.
Has there been a recent update to 3dmark?
Have you checked if you have v-sync or Radeon chill.
Soldato
- Joined
- 14 Aug 2018
- Posts
- 4,009
What surprised you?
I've tested the Steel Nomad benchmark, weirdly the score went up from 7791 (from February when the Timespy was higher) to 7968 (now, that Timespy is lower)
So on Timespy I lost score and Steel Nomad I got some... I don't know... I used to like sorting out this problems, now I just want things to work
Last edited:
Same here... will run that myself later and see what it gives...
Well judging by yours and mine, i[m WAY behind... just don't get it and just removed driver and put new one on and same results with 3 different versions now...
Afaik, Timespy scores have regressed a little bit over the past few drivers, don't know whether it's an AMD problem or 3DMark.
However, Steel Nomad is a far better indicator to test your clocks for stability, it'll crash if you push too far, whereas Timespy will complete but you might not be game stable.
I don't know if makes a difference, but I'm running dual watercooling loops, one for the cpu and one for the gpu as my 9070 xt is watercooled.
CPU is not overclocked.
GPU I'm running: -60mv, 2680 memory (2666 effective), fast timings and +10 on power, it boosts to around 3300mhz and draws 350W during the benchmarks.
You could try checking/changing your clock settings/hpet.
Open a CMD prompt as admin.
Run...
bcdedit
post results here if you want, the settings of interest are normally at the bottom so don't need to post the whole output if you don't want to post it all.
look for these settings...
useplatformclock
useplatformtick
disabledynamictick
To manage these settings...
Force setting on...
bcdedit /set useplatformclock Yes
Force off...
bcdedit /set useplatformclock No
let it default
bcdedit /deletevalue useplatformclock
You can then also test disabling HPET in BIOS and/or disabling the device in device manager.
There is a load of conflicting information about these settings online and I spent a bit of time reading up on it and then testing it on my setup.
Which is a pain as it's a stupid number of combinations.
It's been a while but from what I remember you want to keep HPET enabled in bios and windows and just test the settings above.
I can't find my testing notes or really why I ended up with my settings but I currently use...
HPET in BIOS = enabled
HPET in Windows = enabled.
bcdedit settings....
x2apicpolicy Enable
tscsyncpolicy Enhanced
uselegacyapicmode No
useplatformclock No
useplatformtick No
disabledynamictick No
The first three settings you can ignore for now, they didn't seem to make any difference in my testing but I've posted them if you want to read up on them.
iirc most/all of these can just be deleted and not set as they are pretty much the defaults I think, I just wanted to make sure they where forced.
tl;dr
enable HPET in bios, check it is enabled in device manager
bcdedit
if useplatformclock is set to yes, either...
bcdedit /set useplatformclock No
or...
bcdedit /deletevalue useplatformclock
if it isn't there or is set to no, set it to yes or delete it.
reboot
test
Open a CMD prompt as admin.
Run...
bcdedit
post results here if you want, the settings of interest are normally at the bottom so don't need to post the whole output if you don't want to post it all.
look for these settings...
useplatformclock
useplatformtick
disabledynamictick
To manage these settings...
Force setting on...
bcdedit /set useplatformclock Yes
Force off...
bcdedit /set useplatformclock No
let it default
bcdedit /deletevalue useplatformclock
You can then also test disabling HPET in BIOS and/or disabling the device in device manager.
There is a load of conflicting information about these settings online and I spent a bit of time reading up on it and then testing it on my setup.
Which is a pain as it's a stupid number of combinations.
It's been a while but from what I remember you want to keep HPET enabled in bios and windows and just test the settings above.
I can't find my testing notes or really why I ended up with my settings but I currently use...
HPET in BIOS = enabled
HPET in Windows = enabled.
bcdedit settings....
x2apicpolicy Enable
tscsyncpolicy Enhanced
uselegacyapicmode No
useplatformclock No
useplatformtick No
disabledynamictick No
The first three settings you can ignore for now, they didn't seem to make any difference in my testing but I've posted them if you want to read up on them.
iirc most/all of these can just be deleted and not set as they are pretty much the defaults I think, I just wanted to make sure they where forced.
tl;dr
enable HPET in bios, check it is enabled in device manager
bcdedit
if useplatformclock is set to yes, either...
bcdedit /set useplatformclock No
or...
bcdedit /deletevalue useplatformclock
if it isn't there or is set to no, set it to yes or delete it.
reboot
test
I have a power colour hellhound 9070xt that was purchased from overclockers on 6/3/2025 so is 18 months old, that today decided to melt the pci-e power connector as you can see from the photo's provided. The areas in red are what was damaged the most on the gpu power connector.
I was playing the new cyberpunk 2077 2.5 update this morning and have been for the past few days witrh no issue, and noticed a smell for a couple of seconds that went away then the game crashed "thinking this was the norm for cyberpunk" so rebooted my pc but wouldnt boot up. noticed a light on my motherboard which when googled indicated my gpu wasnt recognised. rebooted a few times just to make sure it wasnt a random thing but still the same error so decided to remove the gpu and "re-sit" it but thats when i noticed the problem when i removed the power cable.
Now i know the nvidia 12vhpwr is well known for this issue but rarley heard of AMD pcie connectors doing this.
What should my next streps be with sorting this out?
I was playing the new cyberpunk 2077 2.5 update this morning and have been for the past few days witrh no issue, and noticed a smell for a couple of seconds that went away then the game crashed "thinking this was the norm for cyberpunk" so rebooted my pc but wouldnt boot up. noticed a light on my motherboard which when googled indicated my gpu wasnt recognised. rebooted a few times just to make sure it wasnt a random thing but still the same error so decided to remove the gpu and "re-sit" it but thats when i noticed the problem when i removed the power cable.
Now i know the nvidia 12vhpwr is well known for this issue but rarley heard of AMD pcie connectors doing this.
What should my next streps be with sorting this out?