In my experience, the 6800 GPU identified as “flaky” does not have enough vdd voltage.
Have a few myself which require a bump in voltages or they drop in hash rates regardless of the rest of the settings.
the cards may consume different power due to manufacturing process sometimes, the silicon lottery counts.
by the way what are the SoC settings for the asrock cards ? i liked the mh/s and consumption and my asorck with this settings but no soc changes consumes around 135w
one of those cards are from AsRock manufacter ?
Dear All,
@frxbg (Моля за помощ!).
I am new and having the following issues, unfortunately my rig restart very often. I have made all changes on everything, everything is new. One of my xfx rx 6800 xt on the way (as it had an issue and seller is sending me new one - I changed new motherboard, new cpu, new psu, and all of them were new, now I am completely stressed and lost). Long story in short. I currently have 2 x xfx rx 6700 xt and when I did amd-info, the result is attached and the questions I have are on the pic, and advices and help shall be appreciated . My rig is still restarting and I need to now first that these GPUs are ok before adding my other 3 xfx rx 6800 xt. Please help.
- is not normal
Can’t see if you have the latest stable kernel with drivers, but you want to start there.
Provided no information on the rig itself, especially risers, power cabling, etc., where most hardware issues live.
6700 thread is here
1 Like
yes. just check my screenshot some posts above RX6800 - Efficient Overclocking - #713 by IckZ
1 Like
I may be out of luck, with exactly same settings on one AsRock RX6800 I never can achieve less than 125w
Thank you!
Could you please tell me how to show what is my current kernel.
The overclock I am using is very stable on NBMiner on my other rig where are only rx 67000 xt and no issues for 30 days (till now).
What it is not normal what you could advice me to check is it hardware issue or only a software issue. I hope that the GPU is not for change and is working one.
Web dashboard for the worker/rig will show you kernel, OS, and drivers in at least (2) places:
Above your GPUs showing over clocking
and below same:
Note
- Kernel in these pictures is 5.10.0-hiveos #83 (latest stable kernel)
- OS version 0.6-212@211130
- AMD drivers 20.40 (5.11.1001) (included in latest stable kernel)
My system is:
PSU: EVGA Supernova 1300 P+
MotherBoard: MSI b450 Gaming Plus Max
CPU: AMD Ryzen 5 3600
@Grea
RAM: Corsair Vengeance LPX 2x8GB
SSD: nvme: Corsair CSSD-F240GBMP510 Force-Series MP510 240GB
Risers: Ubit PCIE Riser 1x to 16x Graphics Expansion
Everything completely new (I had issues with the motherboard - changed with new; issues with CPU - again changed with new., PSU changed with new, and now I am waiting xfx rx 6800 xt becayse it has the following issue “system reboot at post”. All changes were checked on my previous rig that works). I am totally lost with such a big “luck” and I hope that problem 3 is not a hardware but software one.
You want to upgrade at least #72 kernel, but I have not found a reason NOT to upgrade to latest stable #83.
Get into Hive Shell: enter hive-replace -y --stable
Be patient. When it is done, you may have to run disk-expand
That should have been quick. Then synch your password with your rig.conf with: hive-passwd -conf
I have just execute the command: dmesg
and I am stressed about the last rows in red, if it might help you, to help me.
I would not invest my time troubleshooting a version of the kernel/drivers which do not include at least the #72 updates.
1 Like
I am doing exactly what you told me (enter hive-replace -y --stable) waiting for the update and then shall run (disk-expand) and then shall do (ig.conf with: hive-passwd -conf)… currently waiting system to reboot in 20 sec…
Current result is:
and result when I run amd-info is again the same : problem 3 (timeout: the monitored command dumped core)
Overclocks, risers, power cables, usb cables, etc., are where that error frequently arise as you show.
fwiw: 1075 is absolute max memory overclock for a 6700XT, back it down to 1060 and try and again.
Thank you very much for your support this night.
I just purchased new risers. I shall stop the rig, shall change the gpu (2x XFX 6700 XT) with the other two that I have unmouned (2x XFX 6800 XT), shall run amd-info to check does the problem 3 (timeout: the monitored command dumped core) shall still exist.
I hope this issue is from the risers not from the gpu.
Once gain thank you for your supprt, a lot!