Hello there, i have a RTX 2080TI Founders with samung memory on it. after driver loading i have many artefacts, and error 43 in device manager. MATS output:

=== MEMORY ERRORS BY SUBPARTITION ===
SUBPART READ ERRORS WRITE ERRORS UNKNOWN ERRS
------- ----------- ------------ ------------
FBIOA0 0 0 0
FBIOA1 0 1100804 0
FBIOB0 0 765388 0
FBIOB1 0 764190 0
FBIOC0 0 0 0
FBIOC1 0 0 0
FBIOD0 0 0 0
FBIOD1 0 0 0
FBIOE0 0 0 0
FBIOE1 0 0 0
FBIOF0 0 0 0
FBIOF1 0 0 0


so it seems that all 3 memory chips under the gpu are broken, but how is it possible that they are broken at same time? should i replace all 3 or should i only reball it? or can it be the gpu himself that have an issue with memory controller? :(

These memory banks are located near the PCI-E connector and there may be mechanical damage to the contacts under the memory due to the bending of the video card. could drop the computer and the video card bent under its weight.
remove these memory banks and you will probably see the reason))

but do you think that all 3 memory chips a the same time can be broken? or can it be the memory controller?
i have test the card several times...when i bend the card the error count varies on all 3 chips...when i bent the card to the CPU i got more memory erros (but always on 3 chips) and when when i bent the card down i got less memory errors...

but can it be the GPU himselve?

Yes, it may be a GPU, but the probability is more on memory. And it is also easier to remove memory than a GPU.

I second to aleksros's statement, all 3 appear to be right above the PCIe connector and could have suffered from bending.
However, if the PCB got bent, that may include the GPU IC pad disconnection too (of either thermal or impact in nature).
If it was up to me, I would lift, examine, and re-seat one of the memory ICs involved (possibly try swapping with another one) and see what that does. You can try reseating the GPU IC as well given you have the tools and skill to do that.
The worst-case would be the memory controller issue.

i also have RTX2080ti and its the same exact problem 3 memory chips are with errors end all of them is located beneath GPU at PCI slot.. and its Samsung memory so I'm sure its mechanical damage done there.. Don’t have time now to remove memory but in next few days will do it and let you know what a problem was there..

Update: I have tested first with simple reflow with flux, memory A1 and B0 still remained unimpressed, but B1 are now fine - no errors.
So, i have desolder A1 and B0 chips and can see the issue...the solderpad on the memory chip itself are tore off, still stick on solder balls on the graphic card PCB. unbelievable. So memory chips physicalical damaged so i must solder new chips - and - card is now working fine

I dont know what i should do with B1 that was repaired with simple reflow...i think i will change it also, or what do you mean?