System Specifications
Ryzen 7 5800X3D
MSI PRO B550-VC
RTX 3080
32GB 3600MHz RAM
Boot Drive: Patriot Burst 240GB
Storage Drives: 2 Seagate Barracuda x1 1TB and x1 2TB
1 XPG SX8200 Gen3 NVMe
NEW: WD_BLACK SN850X (1TB and 2TB ver. tested)
So I've been doing a lot of troubleshooting and I'll try to summarize it as much as possible but it all started with a purchase of a 2TB SN850X late Dec. and getting it early in Jan. I cloned my 240GB SSD to the NVMe and ran into a KERNEL_INPAGE_DATA_ERROR BSOD. After some debugging it disappeared for a while, 2 weeks exactly in which then the crashes started occurring again, too inconsistently to narrow down to anything. After some testing and research (and becoming a disk drive guru installing CrystalDiskMark and HD Tune Pro) I assumed it was a faulty drive as I would get bad sectors reported in the minidump/event viewer and HD Tune Pro error scans would reveal multiple damaged blocks. I packed up the 2TB NVMe and prepared for an RMA and got a 1TB version from Amazon (I threw out the plastic packaging and newegg doesn't accept returns without the full original packaging so originally I was going to use the 1TB packaging, return the 2TB to newegg and then return the Amazon one after receiving the newegg replacement, but patience ran thin and I was considering just using the 1TB NVMe instead and getting a refund on the 2TB drive.) but this led to further testing after realizing even this new 1TB NVMe was facing issues, and I kind of refused to believe that TWO WD NVMe's would come in with bad sectors on arrival, so I went further with my testing.
XPG 8200 Gen3 NVMe works fine in the M.2_1 slot, as well as the M.2_2 slot. I figured out that the SN850X works fine in the M.2_2 slot, reporting no bad sectors and proper speeds in CrystalDiskMark (capping out at gen3 speeds) but would report speeds lower than 1000MB/s read/write in the M2._1 slot; I've had readings as bad as 50MB/s from CrystalDiskMark. HD Tune Pro would also report damaged blocks. It is only after today, where I found out I actually did NOT update my chipset drivers to the latest version, and after doing so the NVMe was no longer detected in diskpart, WD Dashboard, HD Tune Pro or CrystalDiskInfo (would show up in file explorer and disk management) but I had a Windows Update which solved the detection issue and now the drive is detected, HD Tune Pro full error scan reported no damaged blocks but CrystalDiskMark still repots speeds around 1000MB/s read/write
I have tried changing BIOS settings (PCI_E1 Gen Switch and Chipset Gen Switch from auto to Gen4 and Gen3 respectively, no difference. Also tried changing lanes from auto to x8 x8)
but at this point I can confirm it is NOT a problem with the NVMe drives (the damaged blocks/bad sectors were false flags with something to do with my M.2_1 slot regarding Gen4 NVMe's), but I am completely lost on where to go from here
I don't have another Gen4 compatible system to test the SN850X on, I feel like my options are currently down to just running the Gen4 NVMe on the M.2_2 and putting my Gen3 NVMe back in M.2_1 and running both at Gen3 speeds, but at that point I would have just purchased another gen3 drive, but I don't want to go through the long refund/replacement fiasco with newegg and amazon if the NVMe wasn't faulty at all.
Appreciate any information/tests I could further run to narrow down the problem. I have contacted MSI support and they could not help/support agent forwarded it to the engineering team and they said they'd email me in the next 24 hours (26 hours ago but I never had expectations but will update if I get anything from them)
Ryzen 7 5800X3D
MSI PRO B550-VC
RTX 3080
32GB 3600MHz RAM
Boot Drive: Patriot Burst 240GB
Storage Drives: 2 Seagate Barracuda x1 1TB and x1 2TB
1 XPG SX8200 Gen3 NVMe
NEW: WD_BLACK SN850X (1TB and 2TB ver. tested)
So I've been doing a lot of troubleshooting and I'll try to summarize it as much as possible but it all started with a purchase of a 2TB SN850X late Dec. and getting it early in Jan. I cloned my 240GB SSD to the NVMe and ran into a KERNEL_INPAGE_DATA_ERROR BSOD. After some debugging it disappeared for a while, 2 weeks exactly in which then the crashes started occurring again, too inconsistently to narrow down to anything. After some testing and research (and becoming a disk drive guru installing CrystalDiskMark and HD Tune Pro) I assumed it was a faulty drive as I would get bad sectors reported in the minidump/event viewer and HD Tune Pro error scans would reveal multiple damaged blocks. I packed up the 2TB NVMe and prepared for an RMA and got a 1TB version from Amazon (I threw out the plastic packaging and newegg doesn't accept returns without the full original packaging so originally I was going to use the 1TB packaging, return the 2TB to newegg and then return the Amazon one after receiving the newegg replacement, but patience ran thin and I was considering just using the 1TB NVMe instead and getting a refund on the 2TB drive.) but this led to further testing after realizing even this new 1TB NVMe was facing issues, and I kind of refused to believe that TWO WD NVMe's would come in with bad sectors on arrival, so I went further with my testing.
XPG 8200 Gen3 NVMe works fine in the M.2_1 slot, as well as the M.2_2 slot. I figured out that the SN850X works fine in the M.2_2 slot, reporting no bad sectors and proper speeds in CrystalDiskMark (capping out at gen3 speeds) but would report speeds lower than 1000MB/s read/write in the M2._1 slot; I've had readings as bad as 50MB/s from CrystalDiskMark. HD Tune Pro would also report damaged blocks. It is only after today, where I found out I actually did NOT update my chipset drivers to the latest version, and after doing so the NVMe was no longer detected in diskpart, WD Dashboard, HD Tune Pro or CrystalDiskInfo (would show up in file explorer and disk management) but I had a Windows Update which solved the detection issue and now the drive is detected, HD Tune Pro full error scan reported no damaged blocks but CrystalDiskMark still repots speeds around 1000MB/s read/write
I have tried changing BIOS settings (PCI_E1 Gen Switch and Chipset Gen Switch from auto to Gen4 and Gen3 respectively, no difference. Also tried changing lanes from auto to x8 x8)
but at this point I can confirm it is NOT a problem with the NVMe drives (the damaged blocks/bad sectors were false flags with something to do with my M.2_1 slot regarding Gen4 NVMe's), but I am completely lost on where to go from here
I don't have another Gen4 compatible system to test the SN850X on, I feel like my options are currently down to just running the Gen4 NVMe on the M.2_2 and putting my Gen3 NVMe back in M.2_1 and running both at Gen3 speeds, but at that point I would have just purchased another gen3 drive, but I don't want to go through the long refund/replacement fiasco with newegg and amazon if the NVMe wasn't faulty at all.
Appreciate any information/tests I could further run to narrow down the problem. I have contacted MSI support and they could not help/support agent forwarded it to the engineering team and they said they'd email me in the next 24 hours (26 hours ago but I never had expectations but will update if I get anything from them)