I just helped my brother do this last month. He was BSODing for no reason and had ruled everything out. He was reluctant to check for RAM issues for obvious reasons but I convinced him to check.
He had 4 bad addresses and once he blacklisted them he’s been stable and no BSOD since then. Obviously its better to replace it but this works well enough for now.
You might have some luck reseating the ram after cleaning the contacts with isopropyl for single bit errors
If the errors are in the same address range everytime you can definitely reserve that address range in kernel parameters and nothing will be assigned there, you just lose a small amount of memory (Windows can do this too but I can’t remember the process)
Last ditch effort is to manually adjust your ram timings, take note of what tests your ram fails on because there will be a corresponding setting for certain issues, this is very time intensive though and requires days of memtesting
Can’t you just add a kernel parameter to not use specific memory addresses?
I just helped my brother do this last month. He was BSODing for no reason and had ruled everything out. He was reluctant to check for RAM issues for obvious reasons but I convinced him to check.
He had 4 bad addresses and once he blacklisted them he’s been stable and no BSOD since then. Obviously its better to replace it but this works well enough for now.
It is doable.
🤷
You might have some luck reseating the ram after cleaning the contacts with isopropyl for single bit errors
If the errors are in the same address range everytime you can definitely reserve that address range in kernel parameters and nothing will be assigned there, you just lose a small amount of memory (Windows can do this too but I can’t remember the process)
Last ditch effort is to manually adjust your ram timings, take note of what tests your ram fails on because there will be a corresponding setting for certain issues, this is very time intensive though and requires days of memtesting