Setting up the Samsung 980 Pro leads to crashes in games and displays.
Setting up the Samsung 980 Pro leads to crashes in games and displays.
Last month I purchased a Samsung 980 Pro PCIe 4.0 NVMe M.2 SSD. I transferred my previous Sandisk Sata drive to the new slot using Samsung's migration utility, learned how to transform it into a GUID Partition Table and configured the BIOS to UEFI only for startup. Challenges emerged while running games even if they weren't installed on the device. I own another 1TB HDD where games reside, plus a 500GB HDD containing various files and unnecessary data. Regardless of the title, titles like Fortnite, No Man's Sky, Deep Rock Galactic, Minecraft consistently failed after short play sessions. My monitors would freeze, go dark, and sometimes the game would freeze or crash repeatedly. Discord would also stop working, and Fortnite reported insufficient VRAM for textures. The Event Viewer logged crashes, while Radeon Software indicated a safe mode activation. At one point I encountered a BSoD message reading "KERNEL_MODE_HEAP_CORRUPTION." I restored my previous SSD, rebooted, and wiped the new one. When unplugged it again, the same issues persisted. Even after reinstalling Windows 10 Home 64-bit from the Samsung 980 Pro, problems continued. Samsung's diagnostic tools confirmed the new drive functioned properly. I upgraded my BIOS to the latest Prime B350-PLUS version, as it was still in beta. My MSI RX 470 Armor 4GB drivers were updated to version 23.4.1, with two Windows updates applied recently. I ran memtest86 and passed all checks. I also tried reverting to older driver versions (23.3.2 through 23.2.1), but nothing resolved the problem. Using DDU to remove previous drivers before switching helped, though using outdated drivers sometimes caused monitor glitches. I reset the GPU in Radeon Software after a factory reset, setting memory clock to 1200 MHz (default is 1650 MHz). The CPU stayed within normal temperature ranges—around 40-50°C. I wasn't comfortable with voltage readings or overclocking, so I avoided pushing the system beyond its safe limits. I checked WhoCrashed? but no crash dump was generated, suggesting it wasn't a full system failure. The crashes seemed tied to the graphics card, not the monitor. None of these issues occurred before the M.2 slot was connected. Games like No Man's Sky and Deep Rock Galactic frequently froze or crashed, though not in every session. I spent considerable time trying various fixes and stayed detailed about each step. Thank you for reading my post; please share any insights or suggestions. My setup includes: CPU - Ryzen 5 1600, RAM - Corsair Vengeance LPX 16GB DDR4 3200MHz (DOCP at 2666MHz), Motherboard - PRIME B350-PLUS, GPU - MSI RX470 Armor 4GB OC, Power Supply - be quiet! Pure Power 11 80+ Gold, Storage - Sandisk Plus 120GB (old boot drive in MBR), Samsung 980 Pro PCIe 4.0 NVMe M.2 SSD (attempted as new boot device), 1TB HDD (game files), 500GB HDD (other data). Monitors: ASUS VS228-P 60Hz (HDMI port), Acer SB230 75Hz.**
It seems I can mimic certain problems. When I switch my secondary screen to primary, the games still freeze and both displays turn black, but the main screen works fine. While adjusting settings in Radeon software, the secondary monitor stretches out, but turning off clone makes it go dark again. Discord remains stable, but the Radeon tool triggers a safe mode warning without a restart. Windows settings help temporarily, yet the monitor glitches persist when games crash. It’s unclear how these crashes affect display behavior.
I've dealt with similar issues before and noticed the PSU performed quite well.
It seems the problem stems from using a PCIe 4 drive on a motherboard that only supports PCIe 3.0. This compatibility issue is common, especially since many users face similar crashes. To resolve it, download Samsung Magician from their site and flash the drive’s firmware. If you didn’t mention this, consider updating your system as well.