grirvan
I also am having this issue with this card.
My system:
Intel i9-9900k@3.6GHz, RTX 3080 FTW3 ULTRA 10G-P5-3897-KR, 32 GB [16 GB x2] DDR4-3200 Memory Module - GSKILL Ripjaws V, GIGABYTE Z390 AORUS ULTRA, Thermaltake Toughpower Grand RGB 850W 80 PLUS Gold, Full Modular
I am monitoring gpu temps with Precision 1 and cpu temps with MSI Afterburner and have no high temps while gaming, but having repeatable reboot occurrences while gaming.
No overclock of CPU or GPU.
Should I RMA the card or is a fix coming?
I have even changed my Windows 10 power option from high performance to balanced and it does not help.  

 
Model number of the PSU.
 
The only Thermaltake that we know works is the GF1.

Seasonic is sending me a Prime TX-850 as replacement. I will report back to see if this fixes anything.
 
So far my findings have been pretty consistant: the shutting down --- restarting only seems to happen in heavy ray-tracing games with power slider going over anything above 101%. It is definitely power related.

Will have to provide power supply model number tonight from home, but I am being advised by Thermaltake tech support that since I was previously running a RTX 2080ti with no problems and the only thing that has changed is the new RTX 3080 GPU, it is related to the card and I should RMA. I am able to put my 2080ti back in and it runs fine.

grirvan
Will have to provide power supply model number tonight from home, but I am being advised by Thermaltake tech support that since I was previously running a RTX 2080ti with no problems and the only thing that has changed is the new RTX 3080 GPU, it is related to the card and I should RMA. I am able to put my 2080ti back in and it runs fine.

The 2080Ti is not a 3080.

That is not a helpful response.

3090 gobbles up way more power than 2080Ti ever will.
 
Sounds like lazy customer service to me from Thermaltake. Seasonic acknowledged this problem immediately after I submitted RMA inquiry.

jaxkrabbit
3090 gobbles up way more power than 2080Ti ever will.
 
Sounds like lazy customer service to me from Thermaltake. Seasonic acknowledged this problem immediately after I submitted RMA inquiry.

Yes, not even on the same planet.
 
The transient power spikes are MUCH higher as is the power draw between the 3080 FTW3 Ultra and the 2080Ti.
 
Not even close.

I did look into the EVGA power supply power meter questionnaire.
Upon completion, it recommends a 650 watt max. power supply.
I have 850 watts, so I am thinking I have ample power. 

grirvan
I did look into the EVGA power supply power meter questionnaire.
Upon completion, it recommends a 650 watt max. power supply.
I have 850 watts, so I am thinking I have ample power. 

 
Just forget you saw that, it's useless.

grirvan
I did look into the EVGA power supply power meter questionnaire.
Upon completion, it recommends a 650 watt max. power supply.
I have 850 watts, so I am thinking I have ample power. 

 
3090 and 3080 have power spikes in the duration of milliseconds and THAT can be really high. See Igor's lab's review here
 


 
 
Mind you that is only the GPU. Plug in your CPU and other components, you are looking at a huge spike of power draw. Although most PSU can handle AVERAGE rated power draws, those huge spikes tend to trigger the PSU's surge protection. Which is exactly what is causing all these problems.
 
Older power supply and some newer PSU are simply not built to handle these short spikes.
 
 
Quoting igor himself:
 
Peaks with intervals between 1 and 10 ms can lead to shutdowns in very fast reacting protective circuits (OPP, OCP), especially in multi-rail power supplies, although the average power consumption is still within the standard. For this card, I would therefore calculate at least 460 to 500 watts as a proportion of the total secondary power consumption of the system with a normal OC with up to 400 watts as the power limit as the graphics card load, in order to have enough reserves for the worst case scenario. A short excerpt with higher resolution shows us now the 20 ms measurements (10 μS intervals), as I let them run automatically for value determination: